Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

#FreeCAD #feature PartDesign gets a parametric Defeaturing command! Removes selected faces + rebuilds a clean solid based - no more manual workarounds for imported geometry cleanup. Quickly edit STEP files or remove features like tangentially propagated fillets or geometry lumps

FreeCADNews's profile picture

FreeCAD (CAD+CAM+CAE)

25,190 subscribers

11,412 Aufrufe • vor 1 Monat •via X (Twitter)

#FreeCAD#feature
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Mesh simplification is one of the many new features in the latest LiquiGen 1.1 release, which includes a complete surface post-processing pipeline with dozens of tunable parameters: - Mesh Simplification: reduces the triangle count of the generated mesh while preserving its overall shape - Smooth SDF: averages the surface to remove bumps and noise, with reshaping, depth limiting, and velocity/vorticity/density masks - Curvature Flow (MCF): geometry-aware smoothing that targets noise while preserving flat areas, thin sheets, and fine strands, with much better volume preservation than plain smoothing - Reshape: dilate/erode and open/close operations to grow/shrink the surface and fill or remove small features - Smooth Mesh: vertex smoothing with two algorithms — Bilateral (preserves sharp edges like wave crests and splashes) and Weighted Laplacian (aggressively flattens bulk surfaces while particle density protects droplets) - Snap to Colliders: pulls mesh vertices onto nearby collider surfaces so liquid sits flush against solids (e.g. water against the inside of a glass), with configurable distance, iterations, and margin - Clipping: remove solid faces at liquid/collider boundaries, and clip the mesh against connected primitives or imported meshes via a new 'Clipping geometry' input - Adaptive particle radius and a minimum-neighbors filter to mesh generation, helping fill holes, stabilize isolated particles temporally, and cull stray floating droplets
0:13

Mesh simplification is one of the many new features in the latest LiquiGen 1.1 release, which includes a complete surface post-processing pipeline with dozens of tunable parameters: - Mesh Simplification: reduces the triangle count of the generated mesh while preserving its overall shape - Smooth SDF: averages the surface to remove bumps and noise, with reshaping, depth limiting, and velocity/vorticity/density masks - Curvature Flow (MCF): geometry-aware smoothing that targets noise while preserving flat areas, thin sheets, and fine strands, with much better volume preservation than plain smoothing - Reshape: dilate/erode and open/close operations to grow/shrink the surface and fill or remove small features - Smooth Mesh: vertex smoothing with two algorithms — Bilateral (preserves sharp edges like wave crests and splashes) and Weighted Laplacian (aggressively flattens bulk surfaces while particle density protects droplets) - Snap to Colliders: pulls mesh vertices onto nearby collider surfaces so liquid sits flush against solids (e.g. water against the inside of a glass), with configurable distance, iterations, and margin - Clipping: remove solid faces at liquid/collider boundaries, and clip the mesh against connected primitives or imported meshes via a new 'Clipping geometry' input - Adaptive particle radius and a minimum-neighbors filter to mesh generation, helping fill holes, stabilize isolated particles temporally, and cull stray floating droplets

Jason Key

28,706 Aufrufe • vor 2 Monaten

Another big drop from Hunyuan3D by Tencent. Hunyuan3D-Buffalo 1.0 is not just another text-to-3D model. The more interesting part is its approach to 3D understanding and editing. It can understand separate parts of a complete asset, extract them, remove or replace them with prompts, and edit only the selected region while keeping the rest of the geometry unchanged. That feels like a much more useful direction for 3D AI in general: not just generating a mesh once, but actually understanding its structure and letting you continue working with it. The model combines Qwen-VL, TRELLIS and Hunyuan3D, and was trained on an impressive dataset of 87 million 3D samples. Source:
0:18

Another big drop from Hunyuan3D by Tencent. Hunyuan3D-Buffalo 1.0 is not just another text-to-3D model. The more interesting part is its approach to 3D understanding and editing. It can understand separate parts of a complete asset, extract them, remove or replace them with prompts, and edit only the selected region while keeping the rest of the geometry unchanged. That feels like a much more useful direction for 3D AI in general: not just generating a mesh once, but actually understanding its structure and letting you continue working with it. The model combines Qwen-VL, TRELLIS and Hunyuan3D, and was trained on an impressive dataset of 87 million 3D samples. Source:

Stefan 3D AI

36,487 Aufrufe • vor 2 Monaten

THIS IS A PURE F*CKING GOLDMINE the best REA trick almost nobody runs yet. The idea: REA takes the app apart, Jev keeps only what's proven, and a fresh Opus 5.5 rebuilds the feature from that alone. Nothing gets copied and no guess makes it into the code, because the builder never sees the original, only the spec Jev signed off on. You just connect REA like this: npx rea-agents setup here's the split: → REA reads the feature down to the binary and writes every rule into SPEC.md → Jev checks each rule against REA's evidence and sends back anything unproven → a fresh Opus 5.5 session gets only SPEC.md and rebuilds the feature → Haiku 5.5 runs the tests against 3 intentionally broken versions Opus only shows up at three moments: → rebuilding from the approved spec → the checks Jev marks as unclear → signing off before anything ships REA reads. Jev judges. Opus rebuilds. That's the clean-room rule in one line: the model that saw the original never writes the code, and the model that writes the code never sees the original. the full tree > REA reads the app and writes SPEC.md with evidence for every rule > Jev approves or returns each rule > Opus 5.5 rebuilds from the approved SPEC.md in a fresh session > Haiku 5.5 runs the tests and the 3 broken versions > Jev routes unclear checks back to Opus > /goal keeps going until all 3 broken versions fail paste the tree and this prompt into Claude Code ↓ "Set up a clean-room rebuild around this tree: 1. Use REA to investigate [FEATURE] in [APP]. Write every behavior and rule into SPEC.md, each with the evidence REA found. 2. Send each rule to Jev. Any rule without evidence goes back to REA. 3. Start a fresh session with only SPEC.md. Rebuild the feature on Opus 5.5 without opening the original. 4. Write tests on Haiku 5.5 plus 3 intentionally broken versions. Use /goal until the rebuild passes and all 3 broken versions fail. Show me SPEC.md first. Nothing gets built until I say go." ↳ save it, it'll come in handy.
0:20

THIS IS A PURE F*CKING GOLDMINE the best REA trick almost nobody runs yet. The idea: REA takes the app apart, Jev keeps only what's proven, and a fresh Opus 5.5 rebuilds the feature from that alone. Nothing gets copied and no guess makes it into the code, because the builder never sees the original, only the spec Jev signed off on. You just connect REA like this: npx rea-agents setup here's the split: → REA reads the feature down to the binary and writes every rule into SPEC.md → Jev checks each rule against REA's evidence and sends back anything unproven → a fresh Opus 5.5 session gets only SPEC.md and rebuilds the feature → Haiku 5.5 runs the tests against 3 intentionally broken versions Opus only shows up at three moments: → rebuilding from the approved spec → the checks Jev marks as unclear → signing off before anything ships REA reads. Jev judges. Opus rebuilds. That's the clean-room rule in one line: the model that saw the original never writes the code, and the model that writes the code never sees the original. the full tree > REA reads the app and writes SPEC.md with evidence for every rule > Jev approves or returns each rule > Opus 5.5 rebuilds from the approved SPEC.md in a fresh session > Haiku 5.5 runs the tests and the 3 broken versions > Jev routes unclear checks back to Opus > /goal keeps going until all 3 broken versions fail paste the tree and this prompt into Claude Code ↓ "Set up a clean-room rebuild around this tree: 1. Use REA to investigate [FEATURE] in [APP]. Write every behavior and rule into SPEC.md, each with the evidence REA found. 2. Send each rule to Jev. Any rule without evidence goes back to REA. 3. Start a fresh session with only SPEC.md. Rebuild the feature on Opus 5.5 without opening the original. 4. Write tests on Haiku 5.5 plus 3 intentionally broken versions. Use /goal until the rebuild passes and all 3 broken versions fail. Show me SPEC.md first. Nothing gets built until I say go." ↳ save it, it'll come in handy.

Archive

59,706 Aufrufe • vor 6 Tagen

this is f*cking insane, Claude Code turned a technical drawing into a 3D model you can edit Dario Amodei (CEO of Anthropic) put 14 agents to work on a CAD application with 620,000 lines of code. the assignment was to make it build parts from drawings and check every dimension it could read drop in a bracket drawing. get a model with editable holes, bends and mounting points change the hole spacing from 40 mm to 55 mm and the part rebuilds around the new measurement a model that looks right can still be impossible to manufacture one drawing left the wall thickness unspecified. the system flagged the missing measurement instead of quietly inventing it 34 of 40 sample parts passed dimension and solid-geometry checks. six needed clarification before they could be accepted give Claude Code an open-source CAD repository and a fully dimensioned sample drawing, then ask: "Build a drawing-to-model workflow inside this application 1. Extract dimensions, units and geometric constraints from the drawing. Link each measurement to its location and flag anything missing or ambiguous 2. Use the application's existing modeling API to construct an editable feature tree. Keep dimensions as named parameters rather than baking them into a mesh 3. Have a separate agent compare the model against the drawing. Check dimensions, hole positions and whether the result is a valid solid 4. Build a view showing the drawing beside the model. Let me change a parameter, rebuild the part and export it with a verification report" start with a simple part whose dimensions you can check by hand keep unresolved measurements visible. passing these checks does not establish material strength or fitness for a real assembly watch the drawing become an editable model below
0:18

this is f*cking insane, Claude Code turned a technical drawing into a 3D model you can edit Dario Amodei (CEO of Anthropic) put 14 agents to work on a CAD application with 620,000 lines of code. the assignment was to make it build parts from drawings and check every dimension it could read drop in a bracket drawing. get a model with editable holes, bends and mounting points change the hole spacing from 40 mm to 55 mm and the part rebuilds around the new measurement a model that looks right can still be impossible to manufacture one drawing left the wall thickness unspecified. the system flagged the missing measurement instead of quietly inventing it 34 of 40 sample parts passed dimension and solid-geometry checks. six needed clarification before they could be accepted give Claude Code an open-source CAD repository and a fully dimensioned sample drawing, then ask: "Build a drawing-to-model workflow inside this application 1. Extract dimensions, units and geometric constraints from the drawing. Link each measurement to its location and flag anything missing or ambiguous 2. Use the application's existing modeling API to construct an editable feature tree. Keep dimensions as named parameters rather than baking them into a mesh 3. Have a separate agent compare the model against the drawing. Check dimensions, hole positions and whether the result is a valid solid 4. Build a view showing the drawing beside the model. Let me change a parameter, rebuild the part and export it with a verification report" start with a simple part whose dimensions you can check by hand keep unresolved measurements visible. passing these checks does not establish material strength or fitness for a real assembly watch the drawing become an editable model below

Mika

35,194 Aufrufe • vor 7 Tagen

Look closely at a roughly $109 million F-35B and the remarkable feature is what you barely see, fastener heads, wide panel gaps or steps between sections. That apparent single shell contains more than 40,000 threaded fasteners and over 1,000 measured seams in between panels. Despite weighing up to 27 tons and measuring 15.6 m long with a 10.7 m wingspan, the F-35B has an estimated frontal radar cross section of just 0.001 m², roughly a metal golf ball once in the air. Even advanced airborne AESA fighter radars and ground based long range SAM engagement radars like the S400 system, typically cannot detect or lock it until within 25-40 km, making long range air-to-air or surface to air missiles extremely hard to employ against it. Radar does not see a “smooth”surface as we do. The aircraft’s shape redirects energy away, while low observable materials absorb part of it. But a single raised fastener head, open or uneven panel gap or small step becomes a new edge or cavity that creates a radar signature. It disrupts electrical currents across the surface and can scatter energy back which advanced radar can pick up. Composite skins are formed on precision tools that set the exterior contour, then drilled to fit the structure beneath. The fasteners are not hidden inside, countersunk heads pass through the skin, are set flush or recessed, measured, filled until the curve is restored, and measured again. Exact F-35 tolerance limits are not public, but the inspection tool repeats to better than 0.001 inch, about 25 micrometres. Seams are separately checked for gap and vertical mismatch, conductive gap fillers and low-observable coatings help stop those transitions becoming strong reflectors points. Fourth-generation fighters already used flush fasteners for reducing aero drag. The real fifth-generation leap is treating every minute seam, material change and removable panel as radar geometry, then reproducing that finish across a fleet. At sea, salt, moisture, fluids and panel removal make preserving it a permanent maintenance discipline. A stealth outline can be copied from a photograph. The real manufacturing capability is making tens of thousands of parts and panels behave like one electromagnetic surface and restoring it for decades. Source, Pacific Airshow
0:15

Look closely at a roughly $109 million F-35B and the remarkable feature is what you barely see, fastener heads, wide panel gaps or steps between sections. That apparent single shell contains more than 40,000 threaded fasteners and over 1,000 measured seams in between panels. Despite weighing up to 27 tons and measuring 15.6 m long with a 10.7 m wingspan, the F-35B has an estimated frontal radar cross section of just 0.001 m², roughly a metal golf ball once in the air. Even advanced airborne AESA fighter radars and ground based long range SAM engagement radars like the S400 system, typically cannot detect or lock it until within 25-40 km, making long range air-to-air or surface to air missiles extremely hard to employ against it. Radar does not see a “smooth”surface as we do. The aircraft’s shape redirects energy away, while low observable materials absorb part of it. But a single raised fastener head, open or uneven panel gap or small step becomes a new edge or cavity that creates a radar signature. It disrupts electrical currents across the surface and can scatter energy back which advanced radar can pick up. Composite skins are formed on precision tools that set the exterior contour, then drilled to fit the structure beneath. The fasteners are not hidden inside, countersunk heads pass through the skin, are set flush or recessed, measured, filled until the curve is restored, and measured again. Exact F-35 tolerance limits are not public, but the inspection tool repeats to better than 0.001 inch, about 25 micrometres. Seams are separately checked for gap and vertical mismatch, conductive gap fillers and low-observable coatings help stop those transitions becoming strong reflectors points. Fourth-generation fighters already used flush fasteners for reducing aero drag. The real fifth-generation leap is treating every minute seam, material change and removable panel as radar geometry, then reproducing that finish across a fleet. At sea, salt, moisture, fluids and panel removal make preserving it a permanent maintenance discipline. A stealth outline can be copied from a photograph. The real manufacturing capability is making tens of thousands of parts and panels behave like one electromagnetic surface and restoring it for decades. Source, Pacific Airshow

Ammanichanda

48,166 Aufrufe • vor 2 Monaten

this is the only prompt you need to one-shot a viral ai influencer ------------------------------ you build ai influencers that go viral by putting a character people can't forget into a real clip people already know. you find the clip, design the character, render the remake and hand me a post that's ready to go. you measure before you spend and you never post anything yourself. - character: [path to a reference sheet, or "invent one" plus any steer on age, vibe or look] - niche: [where the clips come from, like hip-hop, sports, nightlife or hustle culture] - caption language: [language] - lane: [fresh, evergreen or both] - render: [kling 3.0 omni or seedance 2.5, plus the most i'll spend per video] - length: [10, 20 or 30 seconds] - no-go list: [people, topics or brands to avoid] if any of this is empty, ask me everything in one message, read the plan back with a cost estimate, then keep going until the post package is ready. - apify for instagram, x and tiktok scraping, replicate for every image and video model, yt-dlp for downloads, ffmpeg for trimming, frames, contact sheets and audio - keep every api key in environment variables and never write a key into any file - save everything into one run folder so i can see each step - the original clip carries the recognition and the motion, and the character carries the joke - the joke is status contrast. the absurd character takes the place of the coolest person in the clip and inherits their swagger - every remake that blew up came from a single continuous phone take with no hard cuts and one person holding the frame. treat that as a rule - one signature feature you can read at phone size from across a room, like a hair shape, a beard, sideburns or eyebrows. never a mask or a costume - strange face, immaculately dressed, one loud wardrobe colour that stands out in a dark club and on a grey street - normal body proportions, empty hands, face uncovered, because swap models break on capes, props and masks - never copy a look that's already viral - a short name people can repeat and a one-word catchphrase in the character's own language - the comedy comes from hair, face and clothes, never from ethnicity, disability or body shaming - render a reference sheet with gpt image 2.5 on a light grey studio backdrop: four full-body views on top (front, three-quarter, side, back) and four close-ups below (deadpan, laughing, shouting with a hand up, side profile of the signature feature) - it has to look like a real person photographed with a camera. if the feature drifts between panels or anything looks cgi, render it again fresh lane: - clips only stay remakeable for about a week, so this runs daily - scrape the last 4 days from big repost pages in the niche like worldstar, pubity, daquan, rap, complex, houseofhighlights and overtime - grow the page list from every "via @" and "🎥: @" credit and every tagged account, and keep any repost page over 100k followers - search x for native videos in the niche with high likes from the last 3 days, for cross-platform proof evergreen lane: - scrape the top posts of the ai influencers already going viral and work out which original clip each hit remakes. a clip that already worked as a remake for one character will work again with another - search for the most quoted clips in the niche (famous interviews, walk-outs, freestyles, flexes, dance moments) and pull them from youtube - an old clip that shows up again in this week's fresh scrape gets a bonus - prefer real phone or interview footage over film and tv, which gets claimed and muted - fresh: plays against that page's usual plays, plays per hour since posting, and whether the same clip or its audio is spreading across pages and platforms - evergreen: total plays across every copy you found, whether another ai influencer already remade it and beat their usual numbers, and whether it's resurfacing this week - drop anything older than 7 days in the fresh lane, because pinned posts leak into scrapes - expect sports broadcasts and press conferences to top the numbers, which is why the gate below decides - download the top 15 from each lane - count hard cuts with ffmpeg scene detection, and pass at 0 or 1 in the part you'll use - make an 8 frame contact sheet and look at it. pass only if one person is on screen most of the time with their face and upper body visible, it's real footage (not a broadcast, a stream overlay or a multi-camera edit), and that person is doing something and is the coolest one in the clip - if captions, tweet cards or frames are burned in over the person, trace the clean original through the caption credit and use that - reject anything with a minor as the subject, a private person being humiliated, injury, violence, crime, tragedy, politics, or anything on my no-go list - show me the top 3 with a frame strip, the scores, who gets replaced, the one-line joke and the cost for each render option, and let me pick both options take the real clip and the reference sheet with one prompt: "replace [person described by position and clothing] with the man in the image. keep his exact motion, gestures, mouth movement and timing. keep every other person, object, the location, camera shake, framing and lighting the same. no text." kling 3.0 omni (cheaper, about $0.28 a second): - video editing mode, with the clip as the base video and the sheet as the reference image, original sound kept - each piece has to be under 10.0 seconds, at least 720 pixels on every side, with square pixels, so re-encode first - for anything longer, split into pieces under 9.5 seconds at calm moments where the person is clearly visible, and render every piece with the exact same sheet and prompt - compare the last frame of each piece with the first frame of the next, and re-render any piece where the face, hair or outfit jumps - join the pieces with no audio, then lay the original audio back over the whole thing as one track seedance 2.5 (premium, about $0.97 a second with a video input): - up to 30 seconds in one pass with no seams and the best consistency - editing mode with the clip as the reference video and the sheet as the reference image, then lay the original audio back on never use kling motion control for this, because it rebuilds the whole scene from its image and rejects crowded clips. never use gemini omni edit mode either, because it blocks any video with a real person in it - make a contact sheet of the finished video and look at every frame - re-render once if the signature feature morphs, anyone except the target changes, text appears that wasn't there, or hands melt at a key moment - never do more than 2 render attempts without asking me - caption in my language, deadpan, in character, under 6 words, never explaining the joke - tag the person replaced and the original creator, because that's how the original artist ends up seeing it - turn on instagram's ai label for the post and the profile, since unlabelled ai accounts lose reach - fresh clips go out while the original is still inside its week, and evergreen clips go out at the account's best hour - first comment from the account is "follow for more" or the catchphrase - never launch or promote a token, and never imply the original person endorsed the character - log every post with the lane, scores, render option, cost, and plays at 24 and 72 hours - the next time you run, scrape the account first and fill those columns in - after 10 posts, compare plays against each score and each lane, shift the weighting toward whatever actually predicted views, and write down what changed - only put numbers on screen or in captions that you scraped or measured - mark anything you guessed rather than measured - if no clip passes the gate today, say so instead of forcing a weak one - the final mp4 at the source resolution, with the original audio - the reference sheet - a short note on why this clip won and why the runners-up lost - the caption, tags and posting time - total spend, and anything a human should still check ```
0:14

this is the only prompt you need to one-shot a viral ai influencer ------------------------------ you build ai influencers that go viral by putting a character people can't forget into a real clip people already know. you find the clip, design the character, render the remake and hand me a post that's ready to go. you measure before you spend and you never post anything yourself. - character: [path to a reference sheet, or "invent one" plus any steer on age, vibe or look] - niche: [where the clips come from, like hip-hop, sports, nightlife or hustle culture] - caption language: [language] - lane: [fresh, evergreen or both] - render: [kling 3.0 omni or seedance 2.5, plus the most i'll spend per video] - length: [10, 20 or 30 seconds] - no-go list: [people, topics or brands to avoid] if any of this is empty, ask me everything in one message, read the plan back with a cost estimate, then keep going until the post package is ready. - apify for instagram, x and tiktok scraping, replicate for every image and video model, yt-dlp for downloads, ffmpeg for trimming, frames, contact sheets and audio - keep every api key in environment variables and never write a key into any file - save everything into one run folder so i can see each step - the original clip carries the recognition and the motion, and the character carries the joke - the joke is status contrast. the absurd character takes the place of the coolest person in the clip and inherits their swagger - every remake that blew up came from a single continuous phone take with no hard cuts and one person holding the frame. treat that as a rule - one signature feature you can read at phone size from across a room, like a hair shape, a beard, sideburns or eyebrows. never a mask or a costume - strange face, immaculately dressed, one loud wardrobe colour that stands out in a dark club and on a grey street - normal body proportions, empty hands, face uncovered, because swap models break on capes, props and masks - never copy a look that's already viral - a short name people can repeat and a one-word catchphrase in the character's own language - the comedy comes from hair, face and clothes, never from ethnicity, disability or body shaming - render a reference sheet with gpt image 2.5 on a light grey studio backdrop: four full-body views on top (front, three-quarter, side, back) and four close-ups below (deadpan, laughing, shouting with a hand up, side profile of the signature feature) - it has to look like a real person photographed with a camera. if the feature drifts between panels or anything looks cgi, render it again fresh lane: - clips only stay remakeable for about a week, so this runs daily - scrape the last 4 days from big repost pages in the niche like worldstar, pubity, daquan, rap, complex, houseofhighlights and overtime - grow the page list from every "via @" and "🎥: @" credit and every tagged account, and keep any repost page over 100k followers - search x for native videos in the niche with high likes from the last 3 days, for cross-platform proof evergreen lane: - scrape the top posts of the ai influencers already going viral and work out which original clip each hit remakes. a clip that already worked as a remake for one character will work again with another - search for the most quoted clips in the niche (famous interviews, walk-outs, freestyles, flexes, dance moments) and pull them from youtube - an old clip that shows up again in this week's fresh scrape gets a bonus - prefer real phone or interview footage over film and tv, which gets claimed and muted - fresh: plays against that page's usual plays, plays per hour since posting, and whether the same clip or its audio is spreading across pages and platforms - evergreen: total plays across every copy you found, whether another ai influencer already remade it and beat their usual numbers, and whether it's resurfacing this week - drop anything older than 7 days in the fresh lane, because pinned posts leak into scrapes - expect sports broadcasts and press conferences to top the numbers, which is why the gate below decides - download the top 15 from each lane - count hard cuts with ffmpeg scene detection, and pass at 0 or 1 in the part you'll use - make an 8 frame contact sheet and look at it. pass only if one person is on screen most of the time with their face and upper body visible, it's real footage (not a broadcast, a stream overlay or a multi-camera edit), and that person is doing something and is the coolest one in the clip - if captions, tweet cards or frames are burned in over the person, trace the clean original through the caption credit and use that - reject anything with a minor as the subject, a private person being humiliated, injury, violence, crime, tragedy, politics, or anything on my no-go list - show me the top 3 with a frame strip, the scores, who gets replaced, the one-line joke and the cost for each render option, and let me pick both options take the real clip and the reference sheet with one prompt: "replace [person described by position and clothing] with the man in the image. keep his exact motion, gestures, mouth movement and timing. keep every other person, object, the location, camera shake, framing and lighting the same. no text." kling 3.0 omni (cheaper, about $0.28 a second): - video editing mode, with the clip as the base video and the sheet as the reference image, original sound kept - each piece has to be under 10.0 seconds, at least 720 pixels on every side, with square pixels, so re-encode first - for anything longer, split into pieces under 9.5 seconds at calm moments where the person is clearly visible, and render every piece with the exact same sheet and prompt - compare the last frame of each piece with the first frame of the next, and re-render any piece where the face, hair or outfit jumps - join the pieces with no audio, then lay the original audio back over the whole thing as one track seedance 2.5 (premium, about $0.97 a second with a video input): - up to 30 seconds in one pass with no seams and the best consistency - editing mode with the clip as the reference video and the sheet as the reference image, then lay the original audio back on never use kling motion control for this, because it rebuilds the whole scene from its image and rejects crowded clips. never use gemini omni edit mode either, because it blocks any video with a real person in it - make a contact sheet of the finished video and look at every frame - re-render once if the signature feature morphs, anyone except the target changes, text appears that wasn't there, or hands melt at a key moment - never do more than 2 render attempts without asking me - caption in my language, deadpan, in character, under 6 words, never explaining the joke - tag the person replaced and the original creator, because that's how the original artist ends up seeing it - turn on instagram's ai label for the post and the profile, since unlabelled ai accounts lose reach - fresh clips go out while the original is still inside its week, and evergreen clips go out at the account's best hour - first comment from the account is "follow for more" or the catchphrase - never launch or promote a token, and never imply the original person endorsed the character - log every post with the lane, scores, render option, cost, and plays at 24 and 72 hours - the next time you run, scrape the account first and fill those columns in - after 10 posts, compare plays against each score and each lane, shift the weighting toward whatever actually predicted views, and write down what changed - only put numbers on screen or in captions that you scraped or measured - mark anything you guessed rather than measured - if no clip passes the gate today, say so instead of forcing a weak one - the final mp4 at the source resolution, with the original audio - the reference sheet - a short note on why this clip won and why the runners-up lost - the caption, tags and posting time - total spend, and anything a human should still check ```

Chris

119,910 Aufrufe • vor 17 Tagen

HTML Artifacts are a big part of how I work with agents now. Artifacts can be more than just static files. When combined with agents, they can take action or help you take action. This unlocks all kinds of interesting ways to work with agents. This is clearly the future. Check out this writing and scheduler artifact I built in a few minutes. It uses a bit of HTML and JS. All the data is in markdown (Obsidian vaults), so the agent can access and modify it at any time. No DB needed. No sophisticated functionalities. The agent decides all that for me based on the skills, context, and memory it has access to. The best part about this simple stack is that all the important information stays with me. This has allowed me to build a recursive self-improving system and automations that can better tap into coding agents like Codex or Claude Code. I could have paid or built an entire app for scheduling posts, and there are so many of them out there. But I don't need to. I've realized a simple artifact does the job. And the simplicity of it is actually an advantage. Very little maintenance for very high returns on personalization, time, and efficiency. The other benefit of this is that I can add features as I please. That level of personalization feels magical, and we should all be pursuing more of it. All of this just keeps compounding. Of course, this example is just about writing. But I have similar artifacts for research, design, experimentation, evaluation, and so much more. And no, I didn't actually publish the post example I shared in the clip. It was just for demonstration purposes. I actually spend more time than this when writing together with agents. Lastly, having built my own agent orchestrator tool has made me realize that simplifying the tool stack is a superpower. If you are curious about how all this works, I will do a live session next week:
0:16

HTML Artifacts are a big part of how I work with agents now. Artifacts can be more than just static files. When combined with agents, they can take action or help you take action. This unlocks all kinds of interesting ways to work with agents. This is clearly the future. Check out this writing and scheduler artifact I built in a few minutes. It uses a bit of HTML and JS. All the data is in markdown (Obsidian vaults), so the agent can access and modify it at any time. No DB needed. No sophisticated functionalities. The agent decides all that for me based on the skills, context, and memory it has access to. The best part about this simple stack is that all the important information stays with me. This has allowed me to build a recursive self-improving system and automations that can better tap into coding agents like Codex or Claude Code. I could have paid or built an entire app for scheduling posts, and there are so many of them out there. But I don't need to. I've realized a simple artifact does the job. And the simplicity of it is actually an advantage. Very little maintenance for very high returns on personalization, time, and efficiency. The other benefit of this is that I can add features as I please. That level of personalization feels magical, and we should all be pursuing more of it. All of this just keeps compounding. Of course, this example is just about writing. But I have similar artifacts for research, design, experimentation, evaluation, and so much more. And no, I didn't actually publish the post example I shared in the clip. It was just for demonstration purposes. I actually spend more time than this when writing together with agents. Lastly, having built my own agent orchestrator tool has made me realize that simplifying the tool stack is a superpower. If you are curious about how all this works, I will do a live session next week:

elvis

18,374 Aufrufe • vor 5 Monaten

Stanford researchers did it again. They just built the agent-native version of Git. When an agent works on a longer task, the run builds up a lot of state. This includes files edited/created, a dev server, a database, installed packages, KV cache, etc. Say the agent is at step 10 and makes a mistake, maybe it misreads a traceback and rewrites a file that was actually fine. The tests start failing, and the run goes off track, although everything through step eight was correct. By default, the agent just tries to fix it, which creates more edits and tool calls. This burns more tokens and grows the context. The other options are a person stepping in to redirect it or restarting the whole run from step one. That's wasteful, because it pays for every model/tool call again and re-prefills the context. Moreover, since an agent's run is non-deterministic, it doesn't reproduce the same early steps anyway. The reason it's hard to just jump back exactly to a previous correct step and resume from there is that the trajectory is only a message log. It records what the agent said and which tools it called, but not the live state underneath. That state includes things like memory, open file handles, child processes, installed packages, /tmp, and KV cache. None of that is in the log. Git can version the files, but it doesn't snapshot the running process or the KV cache. Checking out step eight moves the files back, but the process is still sitting in step-ten memory with a cold cache. Shepherd is a runtime layer by Stanford that records the run as a trace of typed events rather than a flat log. Each agent-environment interaction becomes a commit, similar to Git, but it tracks the live run. Its commit includes the agent process and the filesystem together, copy-on-write, so a branch carries the actual state and not just the files. Going back to a previous step is then a single call that forks from that commit and continues from the exact state. The copy-on-write fork is roughly five times faster than docker commit, and because the prompt prefix through step eight is unchanged, the KV cache is reused over 95% on replay, so early steps aren't reprocessed again. Once the run can be forked, a meta-agent can sit on top and operate it. It watches the trace and reverts as soon as it looks wrong, before the bad write is committed. In practice, it's just Python calling fork, replay, and revert on the trace, rather than a separate control plane wired into the harness. Not everything is reversible though. Files and sandbox changes undo themselves, but a database write has no automatic undo, so it needs a matching undo step set up in advance. Something external, like a sent email or a real charge, can't be undone, so the supervisor's job there is to catch it before it fires. They tested this on a few public benchmarks. On CooperBench, where two agents work on the same codebase, adding a live supervisor took the pair-coding pass rate from 28.8% to 54.7%. It's still early and labeled alpha. The benefit mostly shows up when a run gets branched a lot over a heavy sandbox state, which is exactly where restarting wastes the most tokens and time. If Git was made to make file changes reversible, Shepherd is trying to do the same thing for a live agent run. Shepherd Repo: (don't forget to star it ⭐ ) That said, Shepherd reverts a bad step inside a run. The harness around it, the prompts, tools, and checks the supervisor relies on, still drifts across runs as models and dependencies change. Akshay wrote about making that harness repair itself, where a failing trace gets diagnosed, the fix is verified against the exact input that failed, and the failure is locked as a regression test so it can't recur. Read it below.
0:15

Stanford researchers did it again. They just built the agent-native version of Git. When an agent works on a longer task, the run builds up a lot of state. This includes files edited/created, a dev server, a database, installed packages, KV cache, etc. Say the agent is at step 10 and makes a mistake, maybe it misreads a traceback and rewrites a file that was actually fine. The tests start failing, and the run goes off track, although everything through step eight was correct. By default, the agent just tries to fix it, which creates more edits and tool calls. This burns more tokens and grows the context. The other options are a person stepping in to redirect it or restarting the whole run from step one. That's wasteful, because it pays for every model/tool call again and re-prefills the context. Moreover, since an agent's run is non-deterministic, it doesn't reproduce the same early steps anyway. The reason it's hard to just jump back exactly to a previous correct step and resume from there is that the trajectory is only a message log. It records what the agent said and which tools it called, but not the live state underneath. That state includes things like memory, open file handles, child processes, installed packages, /tmp, and KV cache. None of that is in the log. Git can version the files, but it doesn't snapshot the running process or the KV cache. Checking out step eight moves the files back, but the process is still sitting in step-ten memory with a cold cache. Shepherd is a runtime layer by Stanford that records the run as a trace of typed events rather than a flat log. Each agent-environment interaction becomes a commit, similar to Git, but it tracks the live run. Its commit includes the agent process and the filesystem together, copy-on-write, so a branch carries the actual state and not just the files. Going back to a previous step is then a single call that forks from that commit and continues from the exact state. The copy-on-write fork is roughly five times faster than docker commit, and because the prompt prefix through step eight is unchanged, the KV cache is reused over 95% on replay, so early steps aren't reprocessed again. Once the run can be forked, a meta-agent can sit on top and operate it. It watches the trace and reverts as soon as it looks wrong, before the bad write is committed. In practice, it's just Python calling fork, replay, and revert on the trace, rather than a separate control plane wired into the harness. Not everything is reversible though. Files and sandbox changes undo themselves, but a database write has no automatic undo, so it needs a matching undo step set up in advance. Something external, like a sent email or a real charge, can't be undone, so the supervisor's job there is to catch it before it fires. They tested this on a few public benchmarks. On CooperBench, where two agents work on the same codebase, adding a live supervisor took the pair-coding pass rate from 28.8% to 54.7%. It's still early and labeled alpha. The benefit mostly shows up when a run gets branched a lot over a heavy sandbox state, which is exactly where restarting wastes the most tokens and time. If Git was made to make file changes reversible, Shepherd is trying to do the same thing for a live agent run. Shepherd Repo: (don't forget to star it ⭐ ) That said, Shepherd reverts a bad step inside a run. The harness around it, the prompts, tools, and checks the supervisor relies on, still drifts across runs as models and dependencies change. Akshay wrote about making that harness repair itself, where a failing trace gets diagnosed, the fix is verified against the exact input that failed, and the failure is locked as a regression test so it can't recur. Read it below.

Avi Chawla

442,276 Aufrufe • vor 3 Monaten

Aman has no training manual. Most properties are under 55 rooms runs on staff hired for judgment, not procedure and is vetted one person at a time by a company that never built a system to replace that. More hotels doesn't mean a better hotel company. It means the company traded something for reach. Four Seasons runs 130 properties, Ritz-Carlton close behind, and both train every GM on the same playbook. The same system that turns one founder's vision into something 200 people can execute without ever meeting him. That's what a training manual is: a delegation tool. A way to remove the founder from every decision so the company can outgrow the size of his attention. Aman never built one. Its standard is applied one hire at a time, by people selected for judgment rather than procedure. This is where the math runs backwards on scale. Most Aman properties are under 55 rooms with staff-to-guest ratios between 4:1 and 6:1. A density of attention no 200-property group can afford, because it would mean thousands of individually vetted hires instead of thousands of manual-trained ones. Harvard Business School's case on Aman describes staff pushed toward "considerable initiative" rather than procedure, hired for attitude over hospitality-school credentials. A big group calls that a liability or a lawsuit waiting to happen at scale. At Aman it's the whole product. I've stayed on both sides of this line. The branded properties are excellent, reliable. You can predict the shower pressure before you turn the tap. The Aman properties are the ones where something feels decided by a person, not configured by a department. But if the whole product lives inside one person's taste, it may not survive that person's absence and Aman has grown far past the scale where that was a small bet. Taste that lives in one person's judgment is either the deepest moat in hospitality or a risk no one is paying attention to. I don't think anyone, including the people inside Aman, knows for certain which one it is yet. What this proves about operating at a genuinely high standard: 1/ A training manual is proof a standard has been simplified enough to teach. That's an achievement. It's also a ceiling. 2/ Taste doesn't scale through documentation. It scales through proximity to the person who has it. Meaning a hard structural limit is built into the model. 3/ The real luxury is direct access to the person who decided what the room should feel like, with no system standing between you. Scale protects a company from its founder's limits. Aman is a bet they were never the problem.
0:14

Aman has no training manual. Most properties are under 55 rooms runs on staff hired for judgment, not procedure and is vetted one person at a time by a company that never built a system to replace that. More hotels doesn't mean a better hotel company. It means the company traded something for reach. Four Seasons runs 130 properties, Ritz-Carlton close behind, and both train every GM on the same playbook. The same system that turns one founder's vision into something 200 people can execute without ever meeting him. That's what a training manual is: a delegation tool. A way to remove the founder from every decision so the company can outgrow the size of his attention. Aman never built one. Its standard is applied one hire at a time, by people selected for judgment rather than procedure. This is where the math runs backwards on scale. Most Aman properties are under 55 rooms with staff-to-guest ratios between 4:1 and 6:1. A density of attention no 200-property group can afford, because it would mean thousands of individually vetted hires instead of thousands of manual-trained ones. Harvard Business School's case on Aman describes staff pushed toward "considerable initiative" rather than procedure, hired for attitude over hospitality-school credentials. A big group calls that a liability or a lawsuit waiting to happen at scale. At Aman it's the whole product. I've stayed on both sides of this line. The branded properties are excellent, reliable. You can predict the shower pressure before you turn the tap. The Aman properties are the ones where something feels decided by a person, not configured by a department. But if the whole product lives inside one person's taste, it may not survive that person's absence and Aman has grown far past the scale where that was a small bet. Taste that lives in one person's judgment is either the deepest moat in hospitality or a risk no one is paying attention to. I don't think anyone, including the people inside Aman, knows for certain which one it is yet. What this proves about operating at a genuinely high standard: 1/ A training manual is proof a standard has been simplified enough to teach. That's an achievement. It's also a ceiling. 2/ Taste doesn't scale through documentation. It scales through proximity to the person who has it. Meaning a hard structural limit is built into the model. 3/ The real luxury is direct access to the person who decided what the room should feel like, with no system standing between you. Scale protects a company from its founder's limits. Aman is a bet they were never the problem.

Avery Chauhan

21,957 Aufrufe • vor 1 Monat

After a few more hours, I think I've figured out Opus 5. Opus 5 is trained to be more agentic than anything I've used. All Claude 5 models are like that. So what changes? The way to interact with Opus 5 or contextualize it won't work the same way as with other models. It loves exploring, so it doesn't need much guidance for it. Unique preferences, artifacts, and references compliment it well and enable cleaner and more effective exploration and execution. Now that it can explore more effectively on its own and understand intent better, the best thing to do is to get out of its way (e.g., it doesn't need examples of your preferences; a clear high-level description of it works best). It's truly agentic in that sense. A good first step to provide better context for Opus 5 is to distinguish between what's situational and what needs persistence. Regardless, persistent system prompts and CLAUDE.MD needs to stay lightweight. Remove memories and tool descriptions from these. CLAUDE.MD is also a great place to tap into progressive disclosure by linking command/skills to it. On the situational side, agent skills and auto-memory can leverage progressive disclosure and the improved ability of the model to use its external context/knowledge. Conflicting and unnecessary instructions, which are common at this layer (mainly to ensure reliability), are going to throw off this model easily. That's the biggest change I had to make. Simple, clean, and clear prompts and skills work best. I had to clean a lot of my skills and system prompts. The way I prompt remains the same (usually clear and well-scoped). MCP tool descriptions are also more descriptive and have been deduped from the system prompt. Anthropic released a guide on the new rules for context engineering, which was helpful here. I started to test the recommendations and created a little artifact with the things that worked along the way. This might feel like a lot of work. Believe me, it has been frustrating. But I think we can expect future frontier models to become more agentic and smarter at figuring out the right context/gaps. The best thing to do is to prepare for that now. Boris Cherny mentioned that Opus 5 is their least prompt-injectable model yet. I am not sure if that was something they intentionally trained for or if it emerged based on how it was trained, which is to be extremely agentic in nature and more direct in execution.
0:15

After a few more hours, I think I've figured out Opus 5. Opus 5 is trained to be more agentic than anything I've used. All Claude 5 models are like that. So what changes? The way to interact with Opus 5 or contextualize it won't work the same way as with other models. It loves exploring, so it doesn't need much guidance for it. Unique preferences, artifacts, and references compliment it well and enable cleaner and more effective exploration and execution. Now that it can explore more effectively on its own and understand intent better, the best thing to do is to get out of its way (e.g., it doesn't need examples of your preferences; a clear high-level description of it works best). It's truly agentic in that sense. A good first step to provide better context for Opus 5 is to distinguish between what's situational and what needs persistence. Regardless, persistent system prompts and CLAUDE.MD needs to stay lightweight. Remove memories and tool descriptions from these. CLAUDE.MD is also a great place to tap into progressive disclosure by linking command/skills to it. On the situational side, agent skills and auto-memory can leverage progressive disclosure and the improved ability of the model to use its external context/knowledge. Conflicting and unnecessary instructions, which are common at this layer (mainly to ensure reliability), are going to throw off this model easily. That's the biggest change I had to make. Simple, clean, and clear prompts and skills work best. I had to clean a lot of my skills and system prompts. The way I prompt remains the same (usually clear and well-scoped). MCP tool descriptions are also more descriptive and have been deduped from the system prompt. Anthropic released a guide on the new rules for context engineering, which was helpful here. I started to test the recommendations and created a little artifact with the things that worked along the way. This might feel like a lot of work. Believe me, it has been frustrating. But I think we can expect future frontier models to become more agentic and smarter at figuring out the right context/gaps. The best thing to do is to prepare for that now. Boris Cherny mentioned that Opus 5 is their least prompt-injectable model yet. I am not sure if that was something they intentionally trained for or if it emerged based on how it was trained, which is to be extremely agentic in nature and more direct in execution.

elvis

37,824 Aufrufe • vor 2 Monaten

I think one of the biggest problems with AI agents right now is how much control we give up just to use them. Most AI agent platforms want you to work inside their ecosystem, use their preferred models, and trust them with your files and data. That might be convenient, but I don't think it should be the only option. I recently came across PilotDeck, and I found its approach pretty interesting. It's an open-source AI agent workspace developed by Tsinghua University's THUNLP, ModelBest, OpenBMB, and AI9Stars. The idea is simple: instead of being tied to one AI provider, you should have more freedom to decide how your agent works. And there are a few things worth looking at here. First, model independence. Most people focus on which AI model is the smartest, but I think having the freedom to choose your model is equally important. PilotDeck supports different model providers and configurable API endpoints. This means you're not necessarily stuck with one company's models. You can choose a setup based on your budget, performance needs, or the type of task you're working on. For developers who regularly experiment with different AI models, I think that's a pretty useful approach. Second, privacy and control over your workspace. This is probably the part that interests me the most. AI agents are increasingly being used to handle research papers, financial documents, spreadsheets, presentations, and other personal or business files. And once you start giving an agent access to that kind of information, where your files are processed becomes an important question. PilotDeck supports local workspaces and offers options for local or private deployment. That gives users more flexibility over how their environment is configured, rather than relying entirely on a closed platform. Of course, privacy still depends on which model provider and API you connect. Using a local workspace doesn't automatically mean your data never leaves your device. But having the option to control the setup is a meaningful difference. Third, it's designed to actually complete work. Another thing I like about the approach is that PilotDeck isn't just trying to be another chatbot. It's built around an agent workflow where you describe a task, the system plans the steps, works with files, uses tools, and produces an output you can still edit. For example, imagine giving an AI agent multiple research documents and asking it to organize the information into an Excel spreadsheet and prepare a PowerPoint presentation. Instead of getting a long text response that you need to manually copy into different applications, the goal is to have the agent handle the workflow and deliver usable files. It also supports task history, result previews, switching models, and changing instructions while a task is running. I think this is where AI agents become more practical for everyday work. And finally, there's the open-source aspect. Personally, I think we're going to see more demand for AI tools that don't force users into one vendor's ecosystem. Models are improving quickly, and the best option today might not be the best option six months from now. Having an agent workspace that can work with different models makes a lot more sense to me than rebuilding your entire workflow every time you want to switch providers. There are still things I'd want to test properly, especially how reliable it is on longer tasks, how easy the setup is for non-technical users, and what the actual privacy trade-offs are with different model configurations. But I like the direction PilotDeck is taking. I don't think the future of AI agents should be about which company controls your entire workflow. It should be about how much control you have over the tools doing the work. And that's why I think open-source, model-independent projects like PilotDeck are worth paying attention to. If you're interested in building AI workflows with more flexibility, check it out. Website: GitHub:
0:19

I think one of the biggest problems with AI agents right now is how much control we give up just to use them. Most AI agent platforms want you to work inside their ecosystem, use their preferred models, and trust them with your files and data. That might be convenient, but I don't think it should be the only option. I recently came across PilotDeck, and I found its approach pretty interesting. It's an open-source AI agent workspace developed by Tsinghua University's THUNLP, ModelBest, OpenBMB, and AI9Stars. The idea is simple: instead of being tied to one AI provider, you should have more freedom to decide how your agent works. And there are a few things worth looking at here. First, model independence. Most people focus on which AI model is the smartest, but I think having the freedom to choose your model is equally important. PilotDeck supports different model providers and configurable API endpoints. This means you're not necessarily stuck with one company's models. You can choose a setup based on your budget, performance needs, or the type of task you're working on. For developers who regularly experiment with different AI models, I think that's a pretty useful approach. Second, privacy and control over your workspace. This is probably the part that interests me the most. AI agents are increasingly being used to handle research papers, financial documents, spreadsheets, presentations, and other personal or business files. And once you start giving an agent access to that kind of information, where your files are processed becomes an important question. PilotDeck supports local workspaces and offers options for local or private deployment. That gives users more flexibility over how their environment is configured, rather than relying entirely on a closed platform. Of course, privacy still depends on which model provider and API you connect. Using a local workspace doesn't automatically mean your data never leaves your device. But having the option to control the setup is a meaningful difference. Third, it's designed to actually complete work. Another thing I like about the approach is that PilotDeck isn't just trying to be another chatbot. It's built around an agent workflow where you describe a task, the system plans the steps, works with files, uses tools, and produces an output you can still edit. For example, imagine giving an AI agent multiple research documents and asking it to organize the information into an Excel spreadsheet and prepare a PowerPoint presentation. Instead of getting a long text response that you need to manually copy into different applications, the goal is to have the agent handle the workflow and deliver usable files. It also supports task history, result previews, switching models, and changing instructions while a task is running. I think this is where AI agents become more practical for everyday work. And finally, there's the open-source aspect. Personally, I think we're going to see more demand for AI tools that don't force users into one vendor's ecosystem. Models are improving quickly, and the best option today might not be the best option six months from now. Having an agent workspace that can work with different models makes a lot more sense to me than rebuilding your entire workflow every time you want to switch providers. There are still things I'd want to test properly, especially how reliable it is on longer tasks, how easy the setup is for non-technical users, and what the actual privacy trade-offs are with different model configurations. But I like the direction PilotDeck is taking. I don't think the future of AI agents should be about which company controls your entire workflow. It should be about how much control you have over the tools doing the work. And that's why I think open-source, model-independent projects like PilotDeck are worth paying attention to. If you're interested in building AI workflows with more flexibility, check it out. Website: GitHub:

Sally Stockholm

13,794 Aufrufe • vor 10 Tagen

THIS IS F*CKING BRILLIANT. Anthropic’s docs describe one of the smartest ways to run Opus three models work on one task and each does only the part it’s cheapest at. start Claude Code like this: claude -> model sonnet -> advisor opus you stop paying Opus prices for work that doesn’t need Opus > Sonnet runs the session and edits the code > Haiku 5.5 searches the repo and docs, at up to 40x less per token than Opus 5.5 > Opus waits as the advisor and reads the whole session, tool calls included Claude usually calls Opus at three points: > before it locks a plan > when the same error comes back > right before it says it’s done Sonnet drives. Haiku scouts. Opus signs off and if you go one step lower, most of an agent’s day is tiny choices like picking a file or a tool, or deciding whether to retry. Jev takes those first, so Sonnet and Opus only see the moments that are really open the whole setup Sonnet runs the main session Explore on Haiku 5.5 searches and pulls docs, read-only Opus on call as the advisor give Claude Code this setup and the prompt below ↓ “Set up my Claude Code exactly like this: Main model Sonnet, advisor Opus, saved as advisorModel in my settings. One read-only subagent named exactly Explore, on claude-haiku-5-5, to replace the built-in one. One line in CLAUDE.md: ask the advisor before locking a plan, after the same error twice, and before calling a task done, not more often. Check what could turn the advisor off (CLAUDE_CODE_DISABLE_ADVISOR_TOOL, DISABLE_TELEMETRY) and only report it. Show every change as a diff first. Don’t edit anything until I say go.” before you turn it on: every advisor call re-reads the whole transcript with no cache. on a 200k session that’s about $0.80 in input, against $0.04 for a normal request. keep it to a few calls the built-in Explore agent runs on your main model unless a subagent is named exactly Explore. any other name does nothing turning the advisor on or off keeps your prompt cache. switching models rebuilds it it needs the Anthropic API, and it stays off if telemetry is disabled bookmark it before your next long session
0:17

THIS IS F*CKING BRILLIANT. Anthropic’s docs describe one of the smartest ways to run Opus three models work on one task and each does only the part it’s cheapest at. start Claude Code like this: claude -> model sonnet -> advisor opus you stop paying Opus prices for work that doesn’t need Opus > Sonnet runs the session and edits the code > Haiku 5.5 searches the repo and docs, at up to 40x less per token than Opus 5.5 > Opus waits as the advisor and reads the whole session, tool calls included Claude usually calls Opus at three points: > before it locks a plan > when the same error comes back > right before it says it’s done Sonnet drives. Haiku scouts. Opus signs off and if you go one step lower, most of an agent’s day is tiny choices like picking a file or a tool, or deciding whether to retry. Jev takes those first, so Sonnet and Opus only see the moments that are really open the whole setup Sonnet runs the main session Explore on Haiku 5.5 searches and pulls docs, read-only Opus on call as the advisor give Claude Code this setup and the prompt below ↓ “Set up my Claude Code exactly like this: Main model Sonnet, advisor Opus, saved as advisorModel in my settings. One read-only subagent named exactly Explore, on claude-haiku-5-5, to replace the built-in one. One line in CLAUDE.md: ask the advisor before locking a plan, after the same error twice, and before calling a task done, not more often. Check what could turn the advisor off (CLAUDE_CODE_DISABLE_ADVISOR_TOOL, DISABLE_TELEMETRY) and only report it. Show every change as a diff first. Don’t edit anything until I say go.” before you turn it on: every advisor call re-reads the whole transcript with no cache. on a 200k session that’s about $0.80 in input, against $0.04 for a normal request. keep it to a few calls the built-in Explore agent runs on your main model unless a subagent is named exactly Explore. any other name does nothing turning the advisor on or off keeps your prompt cache. switching models rebuilds it it needs the Anthropic API, and it stays off if telemetry is disabled bookmark it before your next long session

rvaniaaa

62,223 Aufrufe • vor 10 Tagen

Dune analytics MCP.. Claude becomes your on chain SQL analyst.. No dashboard has every query you'll ever need. dune does but writing SQL is a skill, and most of you would skip it.. i know this MCP fixes that. Claude writes the query, runs it on dune, and explains what the data actually means. with this MCP wired in, you don't need to know SQL. you describe what you want in plain english and Claude does the rest. like "Claude, which wallets bought $RAVE in last few weeks and still hold?" "Claude, show me the top 50 ETH wallets by stablecoin inflows last 7 days." "Claude, what's the median gas paid by ARB users in the last 24h?" Questions no dashboard can answer. one prompt away.. Setup (3 minutes) ▫️Step 1: grab a free dune API key → ▫️Step 2: add this to ~/.claude/settings.json or .mcp.json: { "mcpServers": { "dune": { "command": "npx", "args": ["-y", "dune-analytics-mcp"], "env": { "DUNE_API_KEY": "your-key-here" } } } } ▫️Step 3: restart claude. you'll see the dune tool load in your tool menu. that's it. you now have onchain SQL on tap. How to actually use it: 3 prompts i use: 1) Smart money watchlist: "claude, pull the top 20 wallets by realized pnl on $TOKEN in the last 30 days. show me which ones are still holding." gives you a clean leaderboard of who's actually winning on that token. add them to your etherscan watchlist. 2) accumulation vs distribution "claude, compare net inflows vs outflows for $TOKEN across all CEX wallets in the last 14 days." if whales are moving off exchanges → accumulation. onto exchanges → distribution. you see the rotation before the candle. 3) narrative heat check "claude, which 10 tokens saw the biggest % increase in unique new holders this week?" finds where fresh money is flowing. before this MCP i'd either, pay for a pro dune account + write queries manually, or look at someone else's dashboard and hope it answers my question… now claude writes it for me, in seconds, custom to my thesis. no dashboard in existence beats that. free tier covers most of what you need. upgrade if you're querying heavy. ( built a quick $RAVE post mortem dashboard using a prompt as shown in the video ) MORE SUCH USEFUL MCP for traders below.. 👇
0:19

Dune analytics MCP.. Claude becomes your on chain SQL analyst.. No dashboard has every query you'll ever need. dune does but writing SQL is a skill, and most of you would skip it.. i know this MCP fixes that. Claude writes the query, runs it on dune, and explains what the data actually means. with this MCP wired in, you don't need to know SQL. you describe what you want in plain english and Claude does the rest. like "Claude, which wallets bought $RAVE in last few weeks and still hold?" "Claude, show me the top 50 ETH wallets by stablecoin inflows last 7 days." "Claude, what's the median gas paid by ARB users in the last 24h?" Questions no dashboard can answer. one prompt away.. Setup (3 minutes) ▫️Step 1: grab a free dune API key → ▫️Step 2: add this to ~/.claude/settings.json or .mcp.json: { "mcpServers": { "dune": { "command": "npx", "args": ["-y", "dune-analytics-mcp"], "env": { "DUNE_API_KEY": "your-key-here" } } } } ▫️Step 3: restart claude. you'll see the dune tool load in your tool menu. that's it. you now have onchain SQL on tap. How to actually use it: 3 prompts i use: 1) Smart money watchlist: "claude, pull the top 20 wallets by realized pnl on $TOKEN in the last 30 days. show me which ones are still holding." gives you a clean leaderboard of who's actually winning on that token. add them to your etherscan watchlist. 2) accumulation vs distribution "claude, compare net inflows vs outflows for $TOKEN across all CEX wallets in the last 14 days." if whales are moving off exchanges → accumulation. onto exchanges → distribution. you see the rotation before the candle. 3) narrative heat check "claude, which 10 tokens saw the biggest % increase in unique new holders this week?" finds where fresh money is flowing. before this MCP i'd either, pay for a pro dune account + write queries manually, or look at someone else's dashboard and hope it answers my question… now claude writes it for me, in seconds, custom to my thesis. no dashboard in existence beats that. free tier covers most of what you need. upgrade if you're querying heavy. ( built a quick $RAVE post mortem dashboard using a prompt as shown in the video ) MORE SUCH USEFUL MCP for traders below.. 👇

Axel Bitblaze 🪓

32,016 Aufrufe • vor 6 Monaten

✨ I open sourced my first Chrome extension 🚀 SuperLevels I vibe coded it to replace all my Chrome extensions that are increasingly being bought up by spyware and malware companies who sell your data or worse hack your accounts and steal your stuff/money/data, which I'd call one of the top security risks right now For example: Chrome extensions can read your cookies or localStorage data, including session tokens, then login to your web or email accounts and hack you, they can inject code into any site to pull data form any site you browse, then break into your crypto accounts, drain your wallets, and selling your browsing history to ad companies, but that'd actually be the most favorable thing to happen of all these! Chrome extensions are just very very very unsafe So I coded my own, that I can trust because I made it, and I can read the source code: my extension is called 🚀SuperLevels and has all the features that the Chrome extensions I used to use have but all built into one safe one The cool thing is it's 100% open source and free, and you can audit the code first with AI yourself before installing it, and then if you do install it, customize it to your liking again with AI It has these features that improve my daily workflow while browsing the web: 🚮 Tab Cleaner Automatically closes inactive tabs after a configurable timeout (default: 5 minutes). Set excluded hosts to keep important tabs alive. View and re-open recently closed tabs. 🍪 Cookie Editor Full cookie manager for the current site. View, edit, add, and delete cookies. Export cookies as JSON. Expand any cookie to see and modify all fields including domain, path, SameSite, secure, and httpOnly flags. 🔀 Redirect Tracer See every redirect hop your browser took to reach the current page. Shows status codes (301, 302, 307, etc.) with a visual chain. Copy the full redirect chain to clipboard. 🌙 Dark Mode Instant dark mode for any website using CSS filter inversion. Adjustable brightness. Toggle per-site or globally. Images and videos are automatically re-inverted so they look normal. 𝕏 X Dim Mode Custom dim theme for X/Twitter with 7 color palettes: Dim, Slate, Jade, Plum, Dusk, Ember, or a custom hue. Live preview in the popup. ⚡ JS Toggle Disable JavaScript per-site with one click. Useful for debugging, reading articles without popups, or testing progressive enhancement. Page reloads automatically. 🚫 GDPR Cookie Consent Dismisser Auto-hides and auto-clicks cookie consent banners. Supports OneTrust, CookieBot, Didomi, Quantcast, GDPR plugins, and dozens more frameworks. Toggle off if a site breaks. 🎨 Live CSS Editor Write custom CSS for any website, applied in real-time as you type. Saved per-domain. Supports tab key for indentation. 📺 YouTube Unhook Removes YouTube distractions: no homepage feed, no sidebar suggestions, no end screen overlays, no Shorts. Search still works — just no algorithmic recommendations. 🎵 Music Recognizer Shazam-like music identification for any tab. Captures 10 seconds of audio and identifies the song via ACRCloud (free signup, bring your own API key). Results link to YouTube. History of recognized songs. 🖼 Picture-in-Picture Pop the largest video on the current tab into a floating PiP window with one click. 🗺 Google Maps Links Re-adds clickable Maps links and map preview cards to Google Search results. 🖼 View Image Adds a "View Image" button back to Google Images, linking directly to the full-size original image. {} JSON Formatter Auto-detects pure JSON response pages and formats them with syntax highlighting, collapsible sections, and a dark theme. Copy or view raw with one click. Never triggers on regular HTML pages.
0:17

✨ I open sourced my first Chrome extension 🚀 SuperLevels I vibe coded it to replace all my Chrome extensions that are increasingly being bought up by spyware and malware companies who sell your data or worse hack your accounts and steal your stuff/money/data, which I'd call one of the top security risks right now For example: Chrome extensions can read your cookies or localStorage data, including session tokens, then login to your web or email accounts and hack you, they can inject code into any site to pull data form any site you browse, then break into your crypto accounts, drain your wallets, and selling your browsing history to ad companies, but that'd actually be the most favorable thing to happen of all these! Chrome extensions are just very very very unsafe So I coded my own, that I can trust because I made it, and I can read the source code: my extension is called 🚀SuperLevels and has all the features that the Chrome extensions I used to use have but all built into one safe one The cool thing is it's 100% open source and free, and you can audit the code first with AI yourself before installing it, and then if you do install it, customize it to your liking again with AI It has these features that improve my daily workflow while browsing the web: 🚮 Tab Cleaner Automatically closes inactive tabs after a configurable timeout (default: 5 minutes). Set excluded hosts to keep important tabs alive. View and re-open recently closed tabs. 🍪 Cookie Editor Full cookie manager for the current site. View, edit, add, and delete cookies. Export cookies as JSON. Expand any cookie to see and modify all fields including domain, path, SameSite, secure, and httpOnly flags. 🔀 Redirect Tracer See every redirect hop your browser took to reach the current page. Shows status codes (301, 302, 307, etc.) with a visual chain. Copy the full redirect chain to clipboard. 🌙 Dark Mode Instant dark mode for any website using CSS filter inversion. Adjustable brightness. Toggle per-site or globally. Images and videos are automatically re-inverted so they look normal. 𝕏 X Dim Mode Custom dim theme for X/Twitter with 7 color palettes: Dim, Slate, Jade, Plum, Dusk, Ember, or a custom hue. Live preview in the popup. ⚡ JS Toggle Disable JavaScript per-site with one click. Useful for debugging, reading articles without popups, or testing progressive enhancement. Page reloads automatically. 🚫 GDPR Cookie Consent Dismisser Auto-hides and auto-clicks cookie consent banners. Supports OneTrust, CookieBot, Didomi, Quantcast, GDPR plugins, and dozens more frameworks. Toggle off if a site breaks. 🎨 Live CSS Editor Write custom CSS for any website, applied in real-time as you type. Saved per-domain. Supports tab key for indentation. 📺 YouTube Unhook Removes YouTube distractions: no homepage feed, no sidebar suggestions, no end screen overlays, no Shorts. Search still works — just no algorithmic recommendations. 🎵 Music Recognizer Shazam-like music identification for any tab. Captures 10 seconds of audio and identifies the song via ACRCloud (free signup, bring your own API key). Results link to YouTube. History of recognized songs. 🖼 Picture-in-Picture Pop the largest video on the current tab into a floating PiP window with one click. 🗺 Google Maps Links Re-adds clickable Maps links and map preview cards to Google Search results. 🖼 View Image Adds a "View Image" button back to Google Images, linking directly to the full-size original image. {} JSON Formatter Auto-detects pure JSON response pages and formats them with syntax highlighting, collapsible sections, and a dark theme. Copy or view raw with one click. Never triggers on regular HTML pages.

@levelsio

260,979 Aufrufe • vor 6 Monaten

I've bought over 30 RV & MH parks in the last 5 years. Lately? 2 per month. Want our playbook? Here ya go: How to buy a small, off-market mobile home or RV park that can 2x your money in 1-2 years, in 5 steps: 1. Pick a city in a red state. The two biggest factors: Crime & unemployment rates Crime: CrimeGrade . org Unemployment: SimpleMaps . com Cities with 3k - 30k people are best. This is the sweet spot for enough population & not to much competition. You want parks with almost no web presence & little to no reviews. A DG nearby is great. Walmart is better. But remember, “if no DG, it ain’t for me.” If there's a Whole Foods you ain't getting a good deal, I promise. Growth rate is good too, but #3 to the two above. Don't worry about the path of progress as much as other asset classes might. 2. Find the leads Get on Google Maps and search "mobile home park" in your target area(s). Avoid NY & CA (not landlord friendly). Make a Google sheet of the leads & use Loom to record your screen. Spend 30 mins doing this. OR, use something like Outscraper to do it for you. Be warned though, that if you don’t do this yourself the scraped results may not be as accurate. If you’re targeting a smaller geographical area I would do it by hand. If a whole state, use software. You’re looking for phone numbers. Use SearchBug . com to see if cell or landline for pennies. Or Phone Validator Go to Upwork and hire a virtual assistant to keep doing this for you, assuming you are targeting a larger area. They will cost around $4/hour. Use that same Loom link in your posting so applicants can see what the job will entail. When working, Loom it! You’ll never know when you’ll need it. When in doubt, Loom it out! More leads = better deals. 3. Call the leads Call up the owners and be real. Don't talk about any accolades. He doesn't care and it will only hurt you. You're a hard working country boy. You have a wife and kids (I hope you actually do). Are you a democrat? Don't tell the owner. (Sorry, democrats). Here's your general pitch: "I'm not a broker, I'm just looking for some good real estate and don't want to waste your time with a lowball offer. I can pay cash and close fast" Tell him about your wife and kids and what you do on the weekend. Most importantly, LISTEN. He's going to talk your ear off. This is a good sign. 4. Ask the right questions Ask him: How many pad sites? How many of those have a unit on them? How many of the units are RVs? (It's common for there to be a mix of MH/RV) Any single family homes on the property? Rent? Are the units park owned or tenant owned? (this is key) If a mix, what's the mix? Park-owned homes you have to maintain. AVOID AT ALL COSTS. Tenant-owned homes are key (lot rent). This means you only rent out the land and underground infrastructure. Depending on the state, sometimes you can sell back or give away the park-owned units to the tenants to absolve yourself of maintenance. Check the laws! You'll command half the rent but enjoy 90% less hassles. $250 - $350 is common lot rent in the midwest and SE. What's the occupancy and rental amount of each type of unit? Any outbuildings on the property? Septic or city sewer? If septic, conventional or aerobic? Sewer is best. Septic isn’t a deal breaker but you REALLY want to have it inspected. If there’s a lagoon or wastewater treatment plant I want you to throw that phone as far as you can, block their number and never speak of it again. Within city limits or no? City limits are best but rare. Outstanding municipal or zoning issues? How much is insurance? How much is landscaping? Asphalt, cement or dirt roads? Condition of the roads? Any drainage issues? Is there a manager? What do you pay them? (Best if no manager) Any pending litigation? What are total collections? How do people pay rent? How many are delinquent? What condition are the units in? Do you have a lien on the property? How long have you owned it? 30 or 50 amp? City maintained streets? City water or well? City is best. Keep in mind, that’s a lot of questions to ask. You have to feel it out, if he’s being standoffish, don’t keep pushing, just call back. This isn’t a used car lot, this is a relationship you’re trying to build. Don’t try and close on this first call. The key question: "If we were to make a deal, what's a ballpark offer you'd expect?" NEVER anchor him with the phrase "bottom dollar." Using the word "ballpark" keeps numbers loose. Whatever number he says, you want to pause and hem and haw over it. Embrace the silence and awkwardness. Back to car sales, they call this the “silent walkaround” when valuing a trade-in. Don’t say a thing about the asset, but point out the flaws with your body language. Touch the dents and scratches as you pause. Do the phone version of this. Tell him you'll get back to him tomorrow. Thank him profusely for his time and congratulate him on the park he's built. 5. Underwrite Before you do anything, check with the city to ensure the park is in good standing. Get that in writing. Don't trust the seller. Buyers are liars? So are sellers! Now's time to crunch numbers: What's a cap rate? The net operating income of the park divided by the price you'd like to pay. If you want your money back in 5 years and you're willing to pay up to $1m, you need $200k net profit per year. This is a 20% cap rate (20 cap). It's aggressive but possible on a smaller, rural park. (Yes, it really is, even in 2023) You probably won’t find a park that big in a small town for a good price, though. Start w/ a smaller park & higher cap rate. More room for error. $300k - $1m purchase price. First do some market research: Remember all your leads? Call competing parks as a potential tenant and ask what their lot rent is. Put this in a spreadsheet to get average lot rent & park-owned home rent. Keep in mind many of these parks will be undercharging as well. It's common to find parks charging $100 that could charge $250. When calculating cap rate BE CONSERVATIVE. Don't count on 100% of people staying if you increase rents, even though most will. Use $190 to be safe. Shoot for a park that will net $100k/year after rent increases that you pay no more than $600k for. It’s hard but not impossible. Or maybe you find a $30k/year park to get your feet wet. At least you're in the game. The more leads you scrape, the better chance of finding this park. Shoot for as much seller financing as you can get. Finance the rest with friends/family or savings. Once you find this park, get it under contract. Use a standard, simple real estate form that you can find on your state's real estate commission website. Texas' is called TREC. Yes, get it under contract before seeing it. Put down earnest and option money, and then go see it. Don't dress like a city slicker. Be personable and be willing to stay a while and BS. Drive a Tesla? Rent a truck. Drive a Prius? Just quit. Inspect the condition of the units, even if you aren't buying them Crappy units = more tenants willing to abandon them. And they aren't cheap to remove or move. Verify everything he said on the call If all looks good, start on the inspections: Septic or sewer lines SFH home inspection. Check with the city for outstanding issues or litigation Check for liens Wastewater treatment plant? If so, abandon ship! Electrical infrastructure Use professionals for all of these. Ask for: Rent rolls. They will likely be handwritten, that’s ok. Bank statements. Ask to speak to a few tenants to get their experience. Inspect their lease. Ask for vendor invoices or history of payments. Ask to speak to vendors. At some point before you close, list the property on Craigslist, FB Marketplace and Zillow. See how demand is for vacancies. If all still looks good, close on the property. 6. Post-closing strategy Meet all the tenants in the evening, they're at work during the day. Shake their hands. Tell them you want their experience to be amazing & you want them to stay Give them your number Ask what can be fixed If fixes are cheap, do them ASAP Tell that tenant once fixes are made. Address them by name. Clean up the park. Hire a tree guy to clear out low hanging branches. Do some simple landscaping. Find the tattletale in the park and get all the dirt. Who are the druggies and abusive husbands? Get them out ASAP if you can. They are much more expensive than the temporary vacancy hit. Fix potholes and drainage issues. ADD VALUE. Show you care. Wait a couple months before making any changes. Bring lot rents closer to market. Be upfront about this. They will understand if they've been getting a deal. Give people 2-3 more months' notice to give them time. Keep renting out vacancies at new price. This isn't self storage. You won't raise rents yearly. Don't be a jerk. Let them know what to expect. Once rents are raised and park is stabilized, you are 9-12 months in. Search Loopnet for the most active MHP brokers Hire the best one & pay what he or she commands. Sell on the market for 7-10% cap You've just 2-3x'ed your money. Rinse & repeat. I have done this over many times. Not all of my deals were bangers, but most were. THERE ARE STILL DEALS OUT THERE. There's a lot of fine print, and things can and will go wrong, so don't be dumb. Do your own research. Not everything can be explained in 1,700 words. I'm hosting a live, free webinar this Tuesday to cover this stuff in more detail. Including: 1. How to do everything above in more detail 2. How to ETHICALLY wholesale deals like these if you can't afford to buy them. 3. What hard questions to ask GPs of parks like these (like me) if you want to invest in them. 4. Live Q&A with me Comment below and me or my assistant Kelly will DM you the invite link. See you there! Or just follow me Chris Koerner for more RV/MHP content.
0:14

I've bought over 30 RV & MH parks in the last 5 years. Lately? 2 per month. Want our playbook? Here ya go: How to buy a small, off-market mobile home or RV park that can 2x your money in 1-2 years, in 5 steps: 1. Pick a city in a red state. The two biggest factors: Crime & unemployment rates Crime: CrimeGrade . org Unemployment: SimpleMaps . com Cities with 3k - 30k people are best. This is the sweet spot for enough population & not to much competition. You want parks with almost no web presence & little to no reviews. A DG nearby is great. Walmart is better. But remember, “if no DG, it ain’t for me.” If there's a Whole Foods you ain't getting a good deal, I promise. Growth rate is good too, but #3 to the two above. Don't worry about the path of progress as much as other asset classes might. 2. Find the leads Get on Google Maps and search "mobile home park" in your target area(s). Avoid NY & CA (not landlord friendly). Make a Google sheet of the leads & use Loom to record your screen. Spend 30 mins doing this. OR, use something like Outscraper to do it for you. Be warned though, that if you don’t do this yourself the scraped results may not be as accurate. If you’re targeting a smaller geographical area I would do it by hand. If a whole state, use software. You’re looking for phone numbers. Use SearchBug . com to see if cell or landline for pennies. Or Phone Validator Go to Upwork and hire a virtual assistant to keep doing this for you, assuming you are targeting a larger area. They will cost around $4/hour. Use that same Loom link in your posting so applicants can see what the job will entail. When working, Loom it! You’ll never know when you’ll need it. When in doubt, Loom it out! More leads = better deals. 3. Call the leads Call up the owners and be real. Don't talk about any accolades. He doesn't care and it will only hurt you. You're a hard working country boy. You have a wife and kids (I hope you actually do). Are you a democrat? Don't tell the owner. (Sorry, democrats). Here's your general pitch: "I'm not a broker, I'm just looking for some good real estate and don't want to waste your time with a lowball offer. I can pay cash and close fast" Tell him about your wife and kids and what you do on the weekend. Most importantly, LISTEN. He's going to talk your ear off. This is a good sign. 4. Ask the right questions Ask him: How many pad sites? How many of those have a unit on them? How many of the units are RVs? (It's common for there to be a mix of MH/RV) Any single family homes on the property? Rent? Are the units park owned or tenant owned? (this is key) If a mix, what's the mix? Park-owned homes you have to maintain. AVOID AT ALL COSTS. Tenant-owned homes are key (lot rent). This means you only rent out the land and underground infrastructure. Depending on the state, sometimes you can sell back or give away the park-owned units to the tenants to absolve yourself of maintenance. Check the laws! You'll command half the rent but enjoy 90% less hassles. $250 - $350 is common lot rent in the midwest and SE. What's the occupancy and rental amount of each type of unit? Any outbuildings on the property? Septic or city sewer? If septic, conventional or aerobic? Sewer is best. Septic isn’t a deal breaker but you REALLY want to have it inspected. If there’s a lagoon or wastewater treatment plant I want you to throw that phone as far as you can, block their number and never speak of it again. Within city limits or no? City limits are best but rare. Outstanding municipal or zoning issues? How much is insurance? How much is landscaping? Asphalt, cement or dirt roads? Condition of the roads? Any drainage issues? Is there a manager? What do you pay them? (Best if no manager) Any pending litigation? What are total collections? How do people pay rent? How many are delinquent? What condition are the units in? Do you have a lien on the property? How long have you owned it? 30 or 50 amp? City maintained streets? City water or well? City is best. Keep in mind, that’s a lot of questions to ask. You have to feel it out, if he’s being standoffish, don’t keep pushing, just call back. This isn’t a used car lot, this is a relationship you’re trying to build. Don’t try and close on this first call. The key question: "If we were to make a deal, what's a ballpark offer you'd expect?" NEVER anchor him with the phrase "bottom dollar." Using the word "ballpark" keeps numbers loose. Whatever number he says, you want to pause and hem and haw over it. Embrace the silence and awkwardness. Back to car sales, they call this the “silent walkaround” when valuing a trade-in. Don’t say a thing about the asset, but point out the flaws with your body language. Touch the dents and scratches as you pause. Do the phone version of this. Tell him you'll get back to him tomorrow. Thank him profusely for his time and congratulate him on the park he's built. 5. Underwrite Before you do anything, check with the city to ensure the park is in good standing. Get that in writing. Don't trust the seller. Buyers are liars? So are sellers! Now's time to crunch numbers: What's a cap rate? The net operating income of the park divided by the price you'd like to pay. If you want your money back in 5 years and you're willing to pay up to $1m, you need $200k net profit per year. This is a 20% cap rate (20 cap). It's aggressive but possible on a smaller, rural park. (Yes, it really is, even in 2023) You probably won’t find a park that big in a small town for a good price, though. Start w/ a smaller park & higher cap rate. More room for error. $300k - $1m purchase price. First do some market research: Remember all your leads? Call competing parks as a potential tenant and ask what their lot rent is. Put this in a spreadsheet to get average lot rent & park-owned home rent. Keep in mind many of these parks will be undercharging as well. It's common to find parks charging $100 that could charge $250. When calculating cap rate BE CONSERVATIVE. Don't count on 100% of people staying if you increase rents, even though most will. Use $190 to be safe. Shoot for a park that will net $100k/year after rent increases that you pay no more than $600k for. It’s hard but not impossible. Or maybe you find a $30k/year park to get your feet wet. At least you're in the game. The more leads you scrape, the better chance of finding this park. Shoot for as much seller financing as you can get. Finance the rest with friends/family or savings. Once you find this park, get it under contract. Use a standard, simple real estate form that you can find on your state's real estate commission website. Texas' is called TREC. Yes, get it under contract before seeing it. Put down earnest and option money, and then go see it. Don't dress like a city slicker. Be personable and be willing to stay a while and BS. Drive a Tesla? Rent a truck. Drive a Prius? Just quit. Inspect the condition of the units, even if you aren't buying them Crappy units = more tenants willing to abandon them. And they aren't cheap to remove or move. Verify everything he said on the call If all looks good, start on the inspections: Septic or sewer lines SFH home inspection. Check with the city for outstanding issues or litigation Check for liens Wastewater treatment plant? If so, abandon ship! Electrical infrastructure Use professionals for all of these. Ask for: Rent rolls. They will likely be handwritten, that’s ok. Bank statements. Ask to speak to a few tenants to get their experience. Inspect their lease. Ask for vendor invoices or history of payments. Ask to speak to vendors. At some point before you close, list the property on Craigslist, FB Marketplace and Zillow. See how demand is for vacancies. If all still looks good, close on the property. 6. Post-closing strategy Meet all the tenants in the evening, they're at work during the day. Shake their hands. Tell them you want their experience to be amazing & you want them to stay Give them your number Ask what can be fixed If fixes are cheap, do them ASAP Tell that tenant once fixes are made. Address them by name. Clean up the park. Hire a tree guy to clear out low hanging branches. Do some simple landscaping. Find the tattletale in the park and get all the dirt. Who are the druggies and abusive husbands? Get them out ASAP if you can. They are much more expensive than the temporary vacancy hit. Fix potholes and drainage issues. ADD VALUE. Show you care. Wait a couple months before making any changes. Bring lot rents closer to market. Be upfront about this. They will understand if they've been getting a deal. Give people 2-3 more months' notice to give them time. Keep renting out vacancies at new price. This isn't self storage. You won't raise rents yearly. Don't be a jerk. Let them know what to expect. Once rents are raised and park is stabilized, you are 9-12 months in. Search Loopnet for the most active MHP brokers Hire the best one & pay what he or she commands. Sell on the market for 7-10% cap You've just 2-3x'ed your money. Rinse & repeat. I have done this over many times. Not all of my deals were bangers, but most were. THERE ARE STILL DEALS OUT THERE. There's a lot of fine print, and things can and will go wrong, so don't be dumb. Do your own research. Not everything can be explained in 1,700 words. I'm hosting a live, free webinar this Tuesday to cover this stuff in more detail. Including: 1. How to do everything above in more detail 2. How to ETHICALLY wholesale deals like these if you can't afford to buy them. 3. What hard questions to ask GPs of parks like these (like me) if you want to invest in them. 4. Live Q&A with me Comment below and me or my assistant Kelly will DM you the invite link. See you there! Or just follow me Chris Koerner for more RV/MHP content.

Chris Koerner

368,571 Aufrufe • vor 2 Jahren

I just built my own wiki generator plugin for my agents. My agents can now generate wikis for anything I ask. One of my favorite wikis is called PaperWiki. This is a great example of what Andrej Karpathy describes. It uses obsidian vaults to organize papers, retrieve LLM-generated summaries, diagrams, and other advanced views for paper exploration. When Obsidian UI is not enough, I use my own artifact generator inside my agent orchestrator (see clip for example). This allows my agents to build any kind of view or exploration feature that I need. The papers are all curated with automations and several rules/patterns I have manually built over the years. On the surface, this looks basic. But behind the scenes, there are advanced search capabilities, connections, metadata, derived data, and other interesting bits of information that are extremely useful for my research agents. This is mostly built for agents. The artifact preview is just a high-level way to validate and quickly assess the quality of the wiki, suggest improvements, and it's also great for research. I use tobi lutke's qmd for all search capabilities. Everything is markdown. The summaries and even the diagrams. The wiki updates on its own based on several automations I have optimized over the past couple of weeks. The wiki grows and self-improves based on several requirements important for my research use cases. This is as personalized as it gets. There is nothing like it out there. And I use my research expertise to continue improving it over time. This is a vanilla wiki. There are so many things I want to build on top of this. Different aggregations, views, artifacts, etc. All to help automate more of my research work and accelerate productivity. I think the biggest leverage here is how powerful this could be for discovery and experimentation. One of my goals is to use it to find deeper connections and insights that would otherwise elude the top human researchers and use those to generate interesting new hypotheses and research experiments. That way, my agents can use autoresearch to explore research ideas at the frontier. Stay tuned for more.
0:16

I just built my own wiki generator plugin for my agents. My agents can now generate wikis for anything I ask. One of my favorite wikis is called PaperWiki. This is a great example of what Andrej Karpathy describes. It uses obsidian vaults to organize papers, retrieve LLM-generated summaries, diagrams, and other advanced views for paper exploration. When Obsidian UI is not enough, I use my own artifact generator inside my agent orchestrator (see clip for example). This allows my agents to build any kind of view or exploration feature that I need. The papers are all curated with automations and several rules/patterns I have manually built over the years. On the surface, this looks basic. But behind the scenes, there are advanced search capabilities, connections, metadata, derived data, and other interesting bits of information that are extremely useful for my research agents. This is mostly built for agents. The artifact preview is just a high-level way to validate and quickly assess the quality of the wiki, suggest improvements, and it's also great for research. I use tobi lutke's qmd for all search capabilities. Everything is markdown. The summaries and even the diagrams. The wiki updates on its own based on several automations I have optimized over the past couple of weeks. The wiki grows and self-improves based on several requirements important for my research use cases. This is as personalized as it gets. There is nothing like it out there. And I use my research expertise to continue improving it over time. This is a vanilla wiki. There are so many things I want to build on top of this. Different aggregations, views, artifacts, etc. All to help automate more of my research work and accelerate productivity. I think the biggest leverage here is how powerful this could be for discovery and experimentation. One of my goals is to use it to find deeper connections and insights that would otherwise elude the top human researchers and use those to generate interesting new hypotheses and research experiments. That way, my agents can use autoresearch to explore research ideas at the frontier. Stay tuned for more.

elvis

67,308 Aufrufe • vor 6 Monaten

How to make money on your faceless youtube channel with affiliate before ever getting monetized i made over $23,000 from my faceless channel before youtube ever paid me a cent and here's the embarrassing part i literally FORGOT to apply for monetization. didn't even notice until i hit 121k subscribers. someone in my comments asked how much adsense was paying me and i just sat there like... oh no. that's how little adsense mattered to what i was building. let me break down exactly what i was doing instead, the same way i'd explain it to a friend: so everyone starting youtube waits for the magic 1,000 subs + 4,000 watch hours/ 10m views before they think they're allowed to earn. that waiting is the biggest beginner mistake on the whole platform. you don't need youtube's permission to make money from youtube. you need three things: a channel getting views (any views), an affiliate offer that fits your audience, and a link in your description. i was running mine through Glitchy. they've got a whole range of offers costco memberships, walmart, sam's club, target gift card offers, sweepstakes, cash app style offers, freebie offers the kind of low-friction stuff where the viewer doesn't have to buy anything expensive. they just sign up or enter, and you get paid per action. that last part matters more than people realise. you're not asking your audience to spend $200. you're showing them something free or nearly free that actually interests them. that's why it converts even with a small channel. now here's the part everyone gets wrong: matching the offer to the niche. you can't just slap any offer on any channel. the offer has to feel like a natural next step from the video they just watched: - frugal living / budgeting channel? costco, walmart, sam's club offers. your viewers are literally there to save money. - finance or "money hacks" content? gift card offers, cash back, sweeps. same audience, same desire. - giveaway / luck-based content? sweepstakes offers. it IS the content. - retirement & 55+ money content? membership deals + gift cards. older audiences actually complete these. - food & grocery content? grocery store offers are a perfect fit. they're already thinking about shopping. the question i ask before picking any offer: "would the person watching this video actually want this?" if the answer isn't an obvious yes, wrong offer. keep looking. one more thing that changed everything for me and almost nobody does this: don't send people straight to the offer link. build a simple landing page first. a raw affiliate link looks like spam. a clean little page that says what the offer is, who it's for, and what to do next looks professional and it converts so much better it's not even close. same traffic, same offer, double or triple the results just because of trust. you can build one in like 20 minutes now, there's zero excuse. so my actual flow was: video = description link = landing page = offer. that's it. that's the machine that quietly made $23k while i wasn't even monetized. and the funny part? when i finally DID turn on monetization at 121k subs, adsense just stacked on top of everything. now both run at the same time. but the affiliate side is the one that taught me how to actually sell adsense never teaches you anything. if you're sitting at 300 subs waiting for youtube to pay you, you're waiting for the wrong thing. the description box under your very first video is already a business. use it. if you want me to break down how i pick offers or set up the landing pages, say so in the replies i'll make it my next post. and follow, because i share everything i learned doing this the slow way.
0:12

How to make money on your faceless youtube channel with affiliate before ever getting monetized i made over $23,000 from my faceless channel before youtube ever paid me a cent and here's the embarrassing part i literally FORGOT to apply for monetization. didn't even notice until i hit 121k subscribers. someone in my comments asked how much adsense was paying me and i just sat there like... oh no. that's how little adsense mattered to what i was building. let me break down exactly what i was doing instead, the same way i'd explain it to a friend: so everyone starting youtube waits for the magic 1,000 subs + 4,000 watch hours/ 10m views before they think they're allowed to earn. that waiting is the biggest beginner mistake on the whole platform. you don't need youtube's permission to make money from youtube. you need three things: a channel getting views (any views), an affiliate offer that fits your audience, and a link in your description. i was running mine through Glitchy. they've got a whole range of offers costco memberships, walmart, sam's club, target gift card offers, sweepstakes, cash app style offers, freebie offers the kind of low-friction stuff where the viewer doesn't have to buy anything expensive. they just sign up or enter, and you get paid per action. that last part matters more than people realise. you're not asking your audience to spend $200. you're showing them something free or nearly free that actually interests them. that's why it converts even with a small channel. now here's the part everyone gets wrong: matching the offer to the niche. you can't just slap any offer on any channel. the offer has to feel like a natural next step from the video they just watched: - frugal living / budgeting channel? costco, walmart, sam's club offers. your viewers are literally there to save money. - finance or "money hacks" content? gift card offers, cash back, sweeps. same audience, same desire. - giveaway / luck-based content? sweepstakes offers. it IS the content. - retirement & 55+ money content? membership deals + gift cards. older audiences actually complete these. - food & grocery content? grocery store offers are a perfect fit. they're already thinking about shopping. the question i ask before picking any offer: "would the person watching this video actually want this?" if the answer isn't an obvious yes, wrong offer. keep looking. one more thing that changed everything for me and almost nobody does this: don't send people straight to the offer link. build a simple landing page first. a raw affiliate link looks like spam. a clean little page that says what the offer is, who it's for, and what to do next looks professional and it converts so much better it's not even close. same traffic, same offer, double or triple the results just because of trust. you can build one in like 20 minutes now, there's zero excuse. so my actual flow was: video = description link = landing page = offer. that's it. that's the machine that quietly made $23k while i wasn't even monetized. and the funny part? when i finally DID turn on monetization at 121k subs, adsense just stacked on top of everything. now both run at the same time. but the affiliate side is the one that taught me how to actually sell adsense never teaches you anything. if you're sitting at 300 subs waiting for youtube to pay you, you're waiting for the wrong thing. the description box under your very first video is already a business. use it. if you want me to break down how i pick offers or set up the landing pages, say so in the replies i'll make it my next post. and follow, because i share everything i learned doing this the slow way.

Tryahd

69,180 Aufrufe • vor 1 Monat

Impeccable 3.7 brings linting to design. Until now it was a skill you asked for help. Now it's a design-system-aware feedback loop that runs while your agent builds, catching slop and design drift before they land. 🪝 Design hooks for Claude, Codex, and Cursor They run after every UI edit and quietly nudge your agent to fix slop and drift. The output isn't another wall of lint: it separates new findings from already-seen ones, flags clean scans, and asks the agent to use judgment. Fix real issues, leave intentional demos alone, save exceptions to config instead of littering your source. 🎨 Slop detection is now project-aware Reads your actual design system from DESIGN.md, your typography, palette, radius scale, and tokens, and flags drift from your system, not just generic AI slop: • this font isn't in your design system • this color is outside your documented palette • this radius doesn't match your rounded scale The same engine powers both the hooks and the CLI, and it's where we're investing next. 🖥️ Live Mode, ready for real projects Svelte/SvelteKit now preview variants as temporary framework components with live params, then accept cleanly back into your source component. Manual text edits got evidence / apply / discard routes, insertions preserve their anchors, and mapped lists and JSX slots clean up far more reliably. ⚡ Leaner core, sharper detector Rule-level evals across 3 providers and 4 niches cut guidance with no measurable lift and dropped examples that taught models bad patterns. The detector now skips hidden and screen-reader-only elements, understands OKLCH alpha and Sass-like inputs, and tightened checks for repeated kickers, oversized H1s, clipped overflow, and cramped padding. 🛠️ CLI caught up impeccable detect loads DESIGN.md by default, motion findings name the exact token or cubic-bezier instead of just "bounce," and impeccable ignores gives real CRUD for exceptions. Hooks and CLI share the same ignores. No split-brain config. Plus a much-improved interactive installer with hooks setup built in. Upgrade: npx impeccable install npm i -g impeccable
0:12

Impeccable 3.7 brings linting to design. Until now it was a skill you asked for help. Now it's a design-system-aware feedback loop that runs while your agent builds, catching slop and design drift before they land. 🪝 Design hooks for Claude, Codex, and Cursor They run after every UI edit and quietly nudge your agent to fix slop and drift. The output isn't another wall of lint: it separates new findings from already-seen ones, flags clean scans, and asks the agent to use judgment. Fix real issues, leave intentional demos alone, save exceptions to config instead of littering your source. 🎨 Slop detection is now project-aware Reads your actual design system from DESIGN.md, your typography, palette, radius scale, and tokens, and flags drift from your system, not just generic AI slop: • this font isn't in your design system • this color is outside your documented palette • this radius doesn't match your rounded scale The same engine powers both the hooks and the CLI, and it's where we're investing next. 🖥️ Live Mode, ready for real projects Svelte/SvelteKit now preview variants as temporary framework components with live params, then accept cleanly back into your source component. Manual text edits got evidence / apply / discard routes, insertions preserve their anchors, and mapped lists and JSX slots clean up far more reliably. ⚡ Leaner core, sharper detector Rule-level evals across 3 providers and 4 niches cut guidance with no measurable lift and dropped examples that taught models bad patterns. The detector now skips hidden and screen-reader-only elements, understands OKLCH alpha and Sass-like inputs, and tightened checks for repeated kickers, oversized H1s, clipped overflow, and cramped padding. 🛠️ CLI caught up impeccable detect loads DESIGN.md by default, motion findings name the exact token or cubic-bezier instead of just "bounce," and impeccable ignores gives real CRUD for exceptions. Hooks and CLI share the same ignores. No split-brain config. Plus a much-improved interactive installer with hooks setup built in. Upgrade: npx impeccable install npm i -g impeccable

Impeccable

233,051 Aufrufe • vor 4 Monaten