正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

ANTHROPIC JUST EXPOSED WHY MOST MULTI AGENT CODING SYSTEMS FAIL BEFORE WRITING A SINGLE LINE OF CODE the model is not the problem the plan is one bad dependency cut gets copied across six agents then three builders produce the wrong result in parallel spec → planner → three... builders → critic → scribe → PR → back into the spec the planner runs once every agent downstream is forced to inherit that decision builder A owns the API builder B owns the data layer builder C owns the interface none of them reads another lane that is what makes the system fast and what makes a weak plan catastrophic each builder gets its own read write test loop red stays inside the lane green unlocks the critic the critic does not ask whether the code looks good it asks whether the entire plan should exist then the scribe turns the trace into the PR plan hash three diffs test proof critic notes zero summaries written from memory if the PR fails the reason goes back into the spec as a new constraint the next run starts where the last one failed one human step remains approve or send it back five minutes instead of five hours a good plan scales a bad plan multiplies save this before your next agent build ↓show more

elune

1,084 subscribers

34,491 次观看 • 1 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

16 条评论

kadd 的头像

kadd1 个月前

Plan quality is the real multiplier

elune 的头像

elune1 个月前

agre w u

Callan 的头像

Callan1 个月前

bad planning parallelizes failure

elune 的头像

elune1 个月前

yeah

beamnxw ./ 的头像

beamnxw ./1 个月前

thx for ur explanations

elune 的头像

elune1 个月前

🫡🫡

Akbar Shaik 的头像

Akbar Shaik1 个月前

The biggest insight here isn't even the number of agents. It's the dependency graph. Give each builder a clearly defined surface, let them operate independently, keep failures contained, then bring everything together only after verification.

SEO Mastery 的头像

SEO Mastery1 个月前

One bad plan breaks everything downstream

Yasir Prototyper 的头像

Yasir Prototyper1 个月前

Basically supply chain optimization. I work with upstream dependencies daily. One bad plan stalls the whole line. Same principle, different stack.

Tony Tong | Founder | Ancient Systems x AI 的头像

Tony Tong | Founder | Ancient Systems x AI1 个月前

A bad assumption early in the pipeline is the quiet killer. My production app tokenized with NLTK while my training pipeline lemmatized with spaCy, and that mismatch alone created more out-of-vocabulary words at inference than any downstream fix could catch.

Natalie Thompson 的头像

Natalie Thompson1 个月前

straight facts the plan really makes or breaks everything

Arick Goomanovsky 的头像

Arick Goomanovsky1 个月前

Right - the model isn't the problem, the plan is. Which is another way of saying the failure lives in how agents are organized and hand off, not in any single agent. Multi-agent coding breaks at the seams, not the nodes.

Tara Curry 的头像

Tara Curry1 个月前

The success of a multi-agent coding system rests on a well-thought-out plan and effective communication among agents, not just the model itself

غالي الاثمان 的头像

غالي الاثمان1 个月前

bro this is so real, planning is literally everything

Shikhar Sugandhi 的头像

Shikhar Sugandhi1 个月前

Really interesting approach. Keeping agents isolated while feeding failures back into the spec makes a lot of sense

Leo Oliemans | Refinery 的头像

Leo Oliemans | Refinery1 个月前

The dependency graph is part of the safety boundary. I’d want each parallel task to carry the assumptions it depends on, then get rechecked before merge. Otherwise the system can be internally consistent and still build the wrong thing.

相关视频

ANTHROPIC LEAKED A 6-AGENT SETUP THAT TURNS ONE PROMPT INTO A FINISHED PR your name shows up in this process exactly once - at the very end, for 5 minutes. spec → planner → 3 builders → critic → scribe → PR → back into the spec the planner runs once and the whole line inherits its decisions - a bad plan executed by five perfect agents is still a bad result. 3 lanes build in parallel - and they only work because the planner cut the steps so none of them reads another's output. tests loop until green and red goes back to the coder - the critic never sees a broken build. the critic rejects to the plan, not the code - tests ask does it run, the critic asks should it exist. the scribe writes the PR from the trace, not from memory - cut this seat and you get 8 PRs nobody can tell apart. the back edge into the spec is the whole trick - today's failures become tomorrow's constraints without you typing them. one human step in the whole thing: approve or send back - 5 minutes instead of 5 hours. save this and read the full graph engineering course below ↓

ANTHROPIC LEAKED A 6-AGENT SETUP THAT TURNS ONE PROMPT INTO A FINISHED PR your name shows up in this process exactly once - at the very end, for 5 minutes. spec → planner → 3 builders → critic → scribe → PR → back into the spec the planner runs once and the whole line inherits its decisions - a bad plan executed by five perfect agents is still a bad result. 3 lanes build in parallel - and they only work because the planner cut the steps so none of them reads another's output. tests loop until green and red goes back to the coder - the critic never sees a broken build. the critic rejects to the plan, not the code - tests ask does it run, the critic asks should it exist. the scribe writes the PR from the trace, not from memory - cut this seat and you get 8 PRs nobody can tell apart. the back edge into the spec is the whole trick - today's failures become tomorrow's constraints without you typing them. one human step in the whole thing: approve or send back - 5 minutes instead of 5 hours. save this and read the full graph engineering course below ↓

Hanako

164,222 次观看 • 1 个月前

ANTHROPIC LEAKED A 6-AGENT ASSEMBLY LINE WHERE ONE PROMPT GOES IN AND A FINISHED PR COMES OUT your name shows up in this process exactly once - at the very end, for 5 minutes. prompt → planner → coder → tester → critic → scribe → PR Opus 5 plans once - a bad plan executed by five perfect agents is still a bad result. Sonnet 5 runs 3 parallel lanes - and they only work because the planner made the steps independent. tests loop until green and red goes back to the coder - the critic never sees a broken build. Opus 5 as the critic rejects to the plan, not the code - tests ask "does it run", the critic asks "should it exist". the PR description writes itself from the trace, not from memory - cut this seat and you get 8 PRs nobody can tell apart. one human step in the whole thing: approve or send back - 5 minutes instead of 5 hours. save the setup - 2 models, 6 seats, 2 return paths ↓

ANTHROPIC LEAKED A 6-AGENT ASSEMBLY LINE WHERE ONE PROMPT GOES IN AND A FINISHED PR COMES OUT your name shows up in this process exactly once - at the very end, for 5 minutes. prompt → planner → coder → tester → critic → scribe → PR Opus 5 plans once - a bad plan executed by five perfect agents is still a bad result. Sonnet 5 runs 3 parallel lanes - and they only work because the planner made the steps independent. tests loop until green and red goes back to the coder - the critic never sees a broken build. Opus 5 as the critic rejects to the plan, not the code - tests ask "does it run", the critic asks "should it exist". the PR description writes itself from the trace, not from memory - cut this seat and you get 8 PRs nobody can tell apart. one human step in the whole thing: approve or send back - 5 minutes instead of 5 hours. save the setup - 2 models, 6 seats, 2 return paths ↓

Sprytix

39,283 次观看 • 1 个月前

ANTHROPIC LEAKED A SETUP WHERE ONE FILE STARTS 7 LOOPS THAT TURN A PROMPT INTO A WORKING BUSINESS AI engineer drops in 1 file - the other 95% of the work closes automatically without them. spec → tasks → code → tests → docs → release → monitoring → back into the spec you don't write tickets anymore - you write the thing tickets get derived from, and Opus 5 cuts the spec into 6 tasks by itself. a task that fails its test comes back to itself, not to you - and runs a second lap without a single notification. 6 lanes of code run in parallel and none of them wait on you - the critic filters the diff before you ever open it. every new production bug becomes a permanent line in the eval set - the system never steps on it twice. monitoring isn't the end of the line, it's the start of the next spec - those back edges are what make it a graph instead of a conveyor. and one place is left where you say yes - release, 5 minutes instead of 5 hours. save this and paste it into Claude Code - and you get 7 loops that run without you ↓

ANTHROPIC LEAKED A SETUP WHERE ONE FILE STARTS 7 LOOPS THAT TURN A PROMPT INTO A WORKING BUSINESS AI engineer drops in 1 file - the other 95% of the work closes automatically without them. spec → tasks → code → tests → docs → release → monitoring → back into the spec you don't write tickets anymore - you write the thing tickets get derived from, and Opus 5 cuts the spec into 6 tasks by itself. a task that fails its test comes back to itself, not to you - and runs a second lap without a single notification. 6 lanes of code run in parallel and none of them wait on you - the critic filters the diff before you ever open it. every new production bug becomes a permanent line in the eval set - the system never steps on it twice. monitoring isn't the end of the line, it's the start of the next spec - those back edges are what make it a graph instead of a conveyor. and one place is left where you say yes - release, 5 minutes instead of 5 hours. save this and paste it into Claude Code - and you get 7 loops that run without you ↓

Sprytix

66,333 次观看 • 1 个月前

ANTHROPIC LEAKED A $4.4M FILE WHERE 4 AGENTS BUILD A BUSINESS DOING $120K A MONTH IN 20 MINUTES four agents run for $0, and the business on top of them pulls $120,000 a month without a single salary scout → builder → seller → critic → back to the scout the scout finds who buys and at what price, the builder makes the page and the checkout the seller writes the offer and the outreach, the critic decides whether it ships at all Haiku 4.5 searches, two Sonnet 5 build and sell, Opus 5 passes the verdict no agent grades its own work - each one is checked by the next in the ring and the ring closes: the critic checks the scout one agent always gives itself an A, four in a ring never do a kickoff meeting alone takes 60 minutes, here the whole business is done in 20 $120,000 a month is 401 customers at $299 but all four come from one model family, and they share one blind spot the agents catch the errors at the edges, the one down the middle only you catch save this and paste it into Claude Code - let it build you this ring ↓

ANTHROPIC LEAKED A $4.4M FILE WHERE 4 AGENTS BUILD A BUSINESS DOING $120K A MONTH IN 20 MINUTES four agents run for $0, and the business on top of them pulls $120,000 a month without a single salary scout → builder → seller → critic → back to the scout the scout finds who buys and at what price, the builder makes the page and the checkout the seller writes the offer and the outreach, the critic decides whether it ships at all Haiku 4.5 searches, two Sonnet 5 build and sell, Opus 5 passes the verdict no agent grades its own work - each one is checked by the next in the ring and the ring closes: the critic checks the scout one agent always gives itself an A, four in a ring never do a kickoff meeting alone takes 60 minutes, here the whole business is done in 20 $120,000 a month is 401 customers at $299 but all four come from one model family, and they share one blind spot the agents catch the errors at the edges, the one down the middle only you catch save this and paste it into Claude Code - let it build you this ring ↓

Sprytix

43,407 次观看 • 14 天前

ANTHROPIC LEAKED A FILE WHERE 4 AGENTS BUILD A BUSINESS DOING $100K A MONTH IN 60 MINUTES four agents run around the clock, cost $0 and work against a $100,000 target scout → architect → 2 builders → critic → back to root what makes it a team isn't the four agents, it's the three contracts written between them the scout hands the architect who already ships this and what they charge the architect hands the builders a spec with the interfaces named the builders hand the critic a diff and the tests that cover it leave any one of those boundaries vague and the work goes missing exactly there Haiku 4.5 triages cheap, Opus 5 decides expensive, Sonnet 5 builds, Opus 5 judges the most expensive model runs twice: once to plan, once to pass verdict two builders in parallel is two ways to disagree - which is exactly what the gate is for $100,000 is 334 customers at $299, $50,000 is 167, $10,000 is 34 and what you pay isn't even visible on that same scale the cost was never the thing standing in your way save this and paste it into Claude Code - let it write your 3 contracts ↓

ANTHROPIC LEAKED A FILE WHERE 4 AGENTS BUILD A BUSINESS DOING $100K A MONTH IN 60 MINUTES four agents run around the clock, cost $0 and work against a $100,000 target scout → architect → 2 builders → critic → back to root what makes it a team isn't the four agents, it's the three contracts written between them the scout hands the architect who already ships this and what they charge the architect hands the builders a spec with the interfaces named the builders hand the critic a diff and the tests that cover it leave any one of those boundaries vague and the work goes missing exactly there Haiku 4.5 triages cheap, Opus 5 decides expensive, Sonnet 5 builds, Opus 5 judges the most expensive model runs twice: once to plan, once to pass verdict two builders in parallel is two ways to disagree - which is exactly what the gate is for $100,000 is 334 customers at $299, $50,000 is 167, $10,000 is 34 and what you pay isn't even visible on that same scale the cost was never the thing standing in your way save this and paste it into Claude Code - let it write your 3 contracts ↓

Sprytix

146,554 次观看 • 18 天前

ANTHROPIC LEAKED A $4.2M FILE WHERE 4 AGENTS BUILD THE JEV SYSTEM IN 60 MINUTES, DOING $100K A MONTH four agents cost $0, and the system they assemble in an hour generates $100K a month scout → builder → critic → operator → JEV every agent is pinned to one exact model version, and none of them runs on "latest" the scout on Haiku 4.5 reads the market, the builder on Sonnet 5 builds everything, the critic on Opus 5 decides what ships the operator runs Haiku through the batch API at half price - the one agent that never has to hurry the builder takes 59% of the whole hour, because it's the one writing, and writing costs 5x more than reading short answers are cheap answers - and the whole economics rests on that one rule pin the versions in ROOT.md and next month JEV behaves exactly the same point it at "latest" and the model underneath updates quietly, no error, and the answers start to drift $100K a month is 334 customers at $299 four versions, one system - pin them and JEV stays JEV save this and paste it into Claude Code - let it build you these 4 agents ↓

ANTHROPIC LEAKED A $4.2M FILE WHERE 4 AGENTS BUILD THE JEV SYSTEM IN 60 MINUTES, DOING $100K A MONTH four agents cost $0, and the system they assemble in an hour generates $100K a month scout → builder → critic → operator → JEV every agent is pinned to one exact model version, and none of them runs on "latest" the scout on Haiku 4.5 reads the market, the builder on Sonnet 5 builds everything, the critic on Opus 5 decides what ships the operator runs Haiku through the batch API at half price - the one agent that never has to hurry the builder takes 59% of the whole hour, because it's the one writing, and writing costs 5x more than reading short answers are cheap answers - and the whole economics rests on that one rule pin the versions in ROOT.md and next month JEV behaves exactly the same point it at "latest" and the model underneath updates quietly, no error, and the answers start to drift $100K a month is 334 customers at $299 four versions, one system - pin them and JEV stays JEV save this and paste it into Claude Code - let it build you these 4 agents ↓

Sprytix

37,977 次观看 • 13 天前

ANTHROPIC LEAKED A $4.8M FILE WHERE 5 AGENTS BUILD A BUSINESS AIMED AT $120K A MONTH IN 60 MINUTES five agents running at zero hold a business doing $120,000 a month scout → architect → 2 builders → critic → operator all five read one ROOT.md, and you write it exactly once 96 agent-minutes of work fit inside 60 real ones the architect starts before the scout has finished the critic starts before the builders are done none of them got faster - they just stopped standing in a queue and that alone is 1.6x for free there is no human between them, so there is no pause anywhere $120K a month is 400 customers at $299 or 24 clients at $5,000 - arithmetic, not a forecast the hour builds the machine, the offer written in ROOT.md decides the number save this and paste it into Claude Code - let it build you these 5 agents ↓

ANTHROPIC LEAKED A $4.8M FILE WHERE 5 AGENTS BUILD A BUSINESS AIMED AT $120K A MONTH IN 60 MINUTES five agents running at zero hold a business doing $120,000 a month scout → architect → 2 builders → critic → operator all five read one ROOT.md, and you write it exactly once 96 agent-minutes of work fit inside 60 real ones the architect starts before the scout has finished the critic starts before the builders are done none of them got faster - they just stopped standing in a queue and that alone is 1.6x for free there is no human between them, so there is no pause anywhere $120K a month is 400 customers at $299 or 24 clients at $5,000 - arithmetic, not a forecast the hour builds the machine, the offer written in ROOT.md decides the number save this and paste it into Claude Code - let it build you these 5 agents ↓

Sprytix

50,437 次观看 • 21 天前

ANTHROPIC LEAKED A SETUP THEY SPENT OVER 4 YEARS BUILDING WHERE A $30,000/MONTH TEAM GETS REPLACED BY $4 A DAY you're not a solo dev - you're one person carrying work the market prices at $30,000 a month. plan → code ×4 → tests → critic → docs → audit → back to the plan Opus cuts the task into steps and runs twice a day - the priciest model comes out at $1.05, not the biggest line on the bill. Sonnet holds 4 files at once across four lanes for $1.20 - the biggest line, and still the cheapest hour you ever bought. tests run themselves until green for $0.50 - you never see a red build again and never lose an hour to one. the critic is a different model than the one that wrote it - $0.72 for the only second pair of eyes a solo dev gets. $0.22 a day closes the docs you've been postponing for years - the cheapest role on the list turned out to be the one you skipped. the auditor at $0.32 closes the loop back into the planner - cut that edge and you get six assistants instead of a system. the left column is a hiring plan, the right column is a config file. save this and paste it into Claude Code - and you get 6 roles for $4 a day ↓

ANTHROPIC LEAKED A SETUP THEY SPENT OVER 4 YEARS BUILDING WHERE A $30,000/MONTH TEAM GETS REPLACED BY $4 A DAY you're not a solo dev - you're one person carrying work the market prices at $30,000 a month. plan → code ×4 → tests → critic → docs → audit → back to the plan Opus cuts the task into steps and runs twice a day - the priciest model comes out at $1.05, not the biggest line on the bill. Sonnet holds 4 files at once across four lanes for $1.20 - the biggest line, and still the cheapest hour you ever bought. tests run themselves until green for $0.50 - you never see a red build again and never lose an hour to one. the critic is a different model than the one that wrote it - $0.72 for the only second pair of eyes a solo dev gets. $0.22 a day closes the docs you've been postponing for years - the cheapest role on the list turned out to be the one you skipped. the auditor at $0.32 closes the loop back into the planner - cut that edge and you get six assistants instead of a system. the left column is a hiring plan, the right column is a config file. save this and paste it into Claude Code - and you get 6 roles for $4 a day ↓

Sprytix

13,699 次观看 • 1 个月前

ANTHROPIC LEAKED ONE FILE WORTH $3.4M WHERE 7 WORKERS BUILD AN AI COMPANY IN 24 HOURS you hire nobody - you paste in 1 file and get 7 roles across three shifts with zero onboarding. scout → architect → builders ×3 → critic → scribe → night crew → chief → back into the file the scout goes through mail, tickets and news - and has the agenda ready before you sit down at the desk. the architect cuts the work into a task graph and holds the decisions - 3 builders run on parallel branches. the critic checks every diff before you ever open one - you're no longer the first to see the mistake. the scribe closes the docs and the changelog - the thing you postponed for 3 years now happens daily, without you. the night crew takes the long jobs and the research - the ones that never fit inside an 8-hour day. the chief rolls up the metrics, writes the morning report, and is the only one of the 7 allowed to write back into the file. that one back edge is what turns a pile of assistants into a company - tomorrow's rules get written tonight. every shift hands the next one a finished artefact - nothing gets re-derived from scratch. you work 8 hours, the fleet works 16, 7 roles and zero onboarding. save this and paste it into Claude Code - and your company stops idling ↓

ANTHROPIC LEAKED ONE FILE WORTH $3.4M WHERE 7 WORKERS BUILD AN AI COMPANY IN 24 HOURS you hire nobody - you paste in 1 file and get 7 roles across three shifts with zero onboarding. scout → architect → builders ×3 → critic → scribe → night crew → chief → back into the file the scout goes through mail, tickets and news - and has the agenda ready before you sit down at the desk. the architect cuts the work into a task graph and holds the decisions - 3 builders run on parallel branches. the critic checks every diff before you ever open one - you're no longer the first to see the mistake. the scribe closes the docs and the changelog - the thing you postponed for 3 years now happens daily, without you. the night crew takes the long jobs and the research - the ones that never fit inside an 8-hour day. the chief rolls up the metrics, writes the morning report, and is the only one of the 7 allowed to write back into the file. that one back edge is what turns a pile of assistants into a company - tomorrow's rules get written tonight. every shift hands the next one a finished artefact - nothing gets re-derived from scratch. you work 8 hours, the fleet works 16, 7 roles and zero onboarding. save this and paste it into Claude Code - and your company stops idling ↓

Noisy

50,907 次观看 • 1 个月前

ANTHROPIC LEAKED A SETUP WHERE A $1.2M A-YEAR TEAM FITS INTO ONE FILE FOR $4 A DAY you hire nobody - you write one file, and $106,500 of salaries in it turn into $4 of tokens. architect → 4 builders → tests → critic → docs → scout → chief → back into the file the architect runs twice a day and decides everything - the priciest model on the sheet is the smallest line on the bill at $1.05. four builders cost $1.20 a day - and a fifth builder is one more number in the file, not one more quarter of recruiting. the tester asks does it run, the critic asks should it exist at all - $1.22 together for two roles a solo dev never fills. nothing reaches you that hasn't been rejected at least once. the scout and the scribe cost 40 cents - the exact two roles that get cut from every real budget. nobody hires a documentation engineer, and everybody wishes they had one. the chief at 14 cents is the only one who writes back into the file - without it this is a list, with it the list edits itself. hiring the eleventh is a paragraph: no req, no interviews, no ramp - and firing is one command. save this and paste it into Claude Code - and you get 10 roles for the price of one coffee ↓

ANTHROPIC LEAKED A SETUP WHERE A $1.2M A-YEAR TEAM FITS INTO ONE FILE FOR $4 A DAY you hire nobody - you write one file, and $106,500 of salaries in it turn into $4 of tokens. architect → 4 builders → tests → critic → docs → scout → chief → back into the file the architect runs twice a day and decides everything - the priciest model on the sheet is the smallest line on the bill at $1.05. four builders cost $1.20 a day - and a fifth builder is one more number in the file, not one more quarter of recruiting. the tester asks does it run, the critic asks should it exist at all - $1.22 together for two roles a solo dev never fills. nothing reaches you that hasn't been rejected at least once. the scout and the scribe cost 40 cents - the exact two roles that get cut from every real budget. nobody hires a documentation engineer, and everybody wishes they had one. the chief at 14 cents is the only one who writes back into the file - without it this is a list, with it the list edits itself. hiring the eleventh is a paragraph: no req, no interviews, no ramp - and firing is one command. save this and paste it into Claude Code - and you get 10 roles for the price of one coffee ↓

Sprytix

47,679 次观看 • 1 个月前

THIS GUY CONNECTED HIS AI AGENTS TO HIS OBSIDIAN AND BUILT A BRAIN THAT LEARNS ON ITS OWN. HERE'S HOW TO BUILD IT Obsidian is just markdown files sitting in a folder. That turns out to be the perfect memory for an AI agent, because an agent can read and write those files directly. He wired his agents into the vault so they pull context from it, do the work, and write what they learned back. The notes aren't the point. The loop is, and it gets sharper every cycle How to build it: 1. Point an agent at your vault. The fastest way, no plugins, no API keys: open a terminal and run npx obsidian-mcp /path/to/your/vault. That exposes your Obsidian folder to Claude as a tool it can read, search, and write to. Add it to your Claude Code or Cowork config and restart 2. Confirm it can see the brain. Ask it: "list the notes in my vault and summarize what's in them." If it reads them back, the connection is live. Now it starts every task with everything the vault already holds instead of from zero 3. Give each agent one job and a write-back rule. Tell it: "research this, then save what you found as a new note in /brain with links to related notes." One agent researches, one summarizes, one plans. Each writes its output back into the vault 4. Close the loop. Add one line to every agent's instructions: "read /brain before starting, write your result back when done." Now each task leaves the vault richer, and the next run reads that before it works. It compounds instead of resetting 5. You only steer. Review what the brain produces, point it at the next thing. The agents handle the reading, writing, and connecting The edge isn't better notes. It's a brain that feeds itself, so the work gets sharper every cycle instead of starting over Bookmark this

THIS GUY CONNECTED HIS AI AGENTS TO HIS OBSIDIAN AND BUILT A BRAIN THAT LEARNS ON ITS OWN. HERE'S HOW TO BUILD IT Obsidian is just markdown files sitting in a folder. That turns out to be the perfect memory for an AI agent, because an agent can read and write those files directly. He wired his agents into the vault so they pull context from it, do the work, and write what they learned back. The notes aren't the point. The loop is, and it gets sharper every cycle How to build it: 1. Point an agent at your vault. The fastest way, no plugins, no API keys: open a terminal and run npx obsidian-mcp /path/to/your/vault. That exposes your Obsidian folder to Claude as a tool it can read, search, and write to. Add it to your Claude Code or Cowork config and restart 2. Confirm it can see the brain. Ask it: "list the notes in my vault and summarize what's in them." If it reads them back, the connection is live. Now it starts every task with everything the vault already holds instead of from zero 3. Give each agent one job and a write-back rule. Tell it: "research this, then save what you found as a new note in /brain with links to related notes." One agent researches, one summarizes, one plans. Each writes its output back into the vault 4. Close the loop. Add one line to every agent's instructions: "read /brain before starting, write your result back when done." Now each task leaves the vault richer, and the next run reads that before it works. It compounds instead of resetting 5. You only steer. Review what the brain produces, point it at the next thing. The agents handle the reading, writing, and connecting The edge isn't better notes. It's a brain that feeds itself, so the work gets sharper every cycle instead of starting over Bookmark this

Yarchi

58,591 次观看 • 3 个月前

$THIS GUY USED OPUS 4.8 + KIMI K2.6 TO CUT HIS CODING BILL FROM $4,000 TO $700/MO. WITH KIMI RUNNING A 300-AGENT CODING FLOOR, HE STOPPED PAYING CLAUDE TO DO EVERYTHING kimi does the cheap heavy lifting. it can spin up hundreds of agents, push through thousands of steps, write the rough code, expand files, draft tests and handle the repetitive work that burns the most tokens opus 4.8 only comes in where the money is worth spending. first to plan the spec and define the rules, then again to tear apart the output, catch weak logic and flag the bugs that a fast swarm can miss that is what changed the economics. kimi handled the bulk of the volume for a fraction of the price, while opus stayed in the loop as the architect and the reviewer instead of the full-time builder most people still use one model for every step and wonder why their costs explode. this guy split the jobs properly. kimi runs wide and cheap. opus goes deep and skeptical. one builds fast, the other makes sure it should ship the real edge is not finding one perfect model. it is knowing where expensive intelligence actually matters and where cheap parallel output is already enough. that is how a $4,000 workflow turns into a $700 system$

THIS GUY USED OPUS 4.8 + KIMI K2.6 TO CUT HIS CODING BILL FROM $4,000 TO $700/MO. WITH KIMI RUNNING A 300-AGENT CODING FLOOR, HE STOPPED PAYING CLAUDE TO DO EVERYTHING kimi does the cheap heavy lifting. it can spin up hundreds of agents, push through thousands of steps, write the rough code, expand files, draft tests and handle the repetitive work that burns the most tokens opus 4.8 only comes in where the money is worth spending. first to plan the spec and define the rules, then again to tear apart the output, catch weak logic and flag the bugs that a fast swarm can miss that is what changed the economics. kimi handled the bulk of the volume for a fraction of the price, while opus stayed in the loop as the architect and the reviewer instead of the full-time builder most people still use one model for every step and wonder why their costs explode. this guy split the jobs properly. kimi runs wide and cheap. opus goes deep and skeptical. one builds fast, the other makes sure it should ship the real edge is not finding one perfect model. it is knowing where expensive intelligence actually matters and where cheap parallel output is already enough. that is how a $4,000 workflow turns into a $700 system

Gipp 🦅

16,989 次观看 • 4 个月前

ANTHROPIC LEAKED 6 RULES THAT RUN 8 LOOPS AT ONCE - AND ONE ENGINEER DOES THE WORK OF A TEAM one engineer with this system replaces an entire team - and costs the company 8 times less You → Opus 5 plan → Fable 5 ×8 execute → Fable 5 review → you: 5 min Opus 5 takes your intent and breaks it into subtasks - a bad plan multiplied across every loop gives the same number of disasters - the plan is the only thing you read carefully one repo - isolated worktrees for each loop - nobody stands on anyone's file - parallel PRs ship while you do nothing the agent runs the tests and fixes until green PR - a red build never reaches your eyes one CLAUDE.md for the whole fleet - without it you get different opinions from every loop and one merge hell the average engineer closes 3-4 tasks a day - this system closes a full sprint overnight with zero working hours Fable 5 pre-reviews every PR before it reaches you - you spend 5 minutes - your throughput is now measured in diffs read per hour bookmark and paste into Claude Code - your first fleet can start tonight

ANTHROPIC LEAKED 6 RULES THAT RUN 8 LOOPS AT ONCE - AND ONE ENGINEER DOES THE WORK OF A TEAM one engineer with this system replaces an entire team - and costs the company 8 times less You → Opus 5 plan → Fable 5 ×8 execute → Fable 5 review → you: 5 min Opus 5 takes your intent and breaks it into subtasks - a bad plan multiplied across every loop gives the same number of disasters - the plan is the only thing you read carefully one repo - isolated worktrees for each loop - nobody stands on anyone's file - parallel PRs ship while you do nothing the agent runs the tests and fixes until green PR - a red build never reaches your eyes one CLAUDE.md for the whole fleet - without it you get different opinions from every loop and one merge hell the average engineer closes 3-4 tasks a day - this system closes a full sprint overnight with zero working hours Fable 5 pre-reviews every PR before it reaches you - you spend 5 minutes - your throughput is now measured in diffs read per hour bookmark and paste into Claude Code - your first fleet can start tonight

Sprytix

37,904 次观看 • 2 个月前

ANTHROPIC LEAKED A FILE WHERE 8 ENGINEERS COST $2.40 A DAY INSTEAD OF $120,000 A MONTH Microsoft and Google already run this - the same 8 seats cost them $73 a month. the ring turns once a day, 3 shifts hand off the work, zero meetings. the scout has tomorrow's agenda ready before you sit down. the architect splits the work, 2 builders run parallel branches. contracts get written before the workers - skip that and every worker guesses what the last one meant. with contracts all three read one shape: what goes in, what comes out. 3 branches at once means you pay the slowest lane, not the sum of three. nothing reaches you unrejected: the cheap model asks does it run, the expensive one asks should it merge. a gate that never rejects anything is a status badge, not a gate. the night crew takes the long jobs that never fit a working day. 1 role out of 8 writes back into the file - cut that edge and the rest repeat monday for a year. save this and paste it into Claude Code - you write the file, the file runs the company ↓

ANTHROPIC LEAKED A FILE WHERE 8 ENGINEERS COST $2.40 A DAY INSTEAD OF $120,000 A MONTH Microsoft and Google already run this - the same 8 seats cost them $73 a month. the ring turns once a day, 3 shifts hand off the work, zero meetings. the scout has tomorrow's agenda ready before you sit down. the architect splits the work, 2 builders run parallel branches. contracts get written before the workers - skip that and every worker guesses what the last one meant. with contracts all three read one shape: what goes in, what comes out. 3 branches at once means you pay the slowest lane, not the sum of three. nothing reaches you unrejected: the cheap model asks does it run, the expensive one asks should it merge. a gate that never rejects anything is a status badge, not a gate. the night crew takes the long jobs that never fit a working day. 1 role out of 8 writes back into the file - cut that edge and the rest repeat monday for a year. save this and paste it into Claude Code - you write the file, the file runs the company ↓

Sprytix

699,444 次观看 • 1 个月前

ANTHROPIC LEAKED A FILE WHERE 5 AGENTS WITH JEV BUILD A $1.2M A YEAR SYSTEM IN 30 MINUTES you pay $3.70 and walk away with an asset that can do $1.2M a year repo → data model → API → UI → evals no layer starts until the one under it is closed, and that edge is what turns five scripts into a system the operator on Haiku 4.5 brings up the repo and CI, the architect on Opus 5 lays down the data model and migrations two builders on Sonnet 5 run API and UI in parallel, and the reviewer on Opus 5.5 can keep finished code out of prod the evals were sitting in the file before the first line of code existed at minute zero there was not a single file, by the end it is a working product with payments and tests five agents cover in half an hour what a human team bills for a quarter JEV hands one line between layers instead of the whole history, so the next agent never re-reads the project and the run costs $3.70 instead of $370 the build takes 30 minutes, the first customers show up weeks later save this and paste it into Claude Code - let it assemble those 5 layers for you ↓

ANTHROPIC LEAKED A FILE WHERE 5 AGENTS WITH JEV BUILD A $1.2M A YEAR SYSTEM IN 30 MINUTES you pay $3.70 and walk away with an asset that can do $1.2M a year repo → data model → API → UI → evals no layer starts until the one under it is closed, and that edge is what turns five scripts into a system the operator on Haiku 4.5 brings up the repo and CI, the architect on Opus 5 lays down the data model and migrations two builders on Sonnet 5 run API and UI in parallel, and the reviewer on Opus 5.5 can keep finished code out of prod the evals were sitting in the file before the first line of code existed at minute zero there was not a single file, by the end it is a working product with payments and tests five agents cover in half an hour what a human team bills for a quarter JEV hands one line between layers instead of the whole history, so the next agent never re-reads the project and the run costs $3.70 instead of $370 the build takes 30 minutes, the first customers show up weeks later save this and paste it into Claude Code - let it assemble those 5 layers for you ↓

Sprytix

14,208 次观看 • 9 天前

Godfather of agents showed how he created the loop engineering that everyone now uses: as PhD student, he wrote the formula every AI agent now runs on - his formula taught agents to self-improve here's the formula: step 1 → stop making the model just think OR just act - make it alternate: think → act → observe → think again step 2 → let it reason out loud before every action - the plan guides the next move, not just the output step 3 → feed the result of each action back in - the model reads what happened and adjusts, live step 4 → when it fails, don't restart make it reflect on WHY it failed, in words, then retry with that memory step 5 → store those reflections - the agent gets better across attempts without any retraining before OpenAI, before he ran Tencent's AI - he was a grad student who quietly wrote the blueprint every agent now follows - now give 90-min course watch & bookmark - then read how the agent systems actually works ↓

Godfather of agents showed how he created the loop engineering that everyone now uses: as PhD student, he wrote the formula every AI agent now runs on - his formula taught agents to self-improve here's the formula: step 1 → stop making the model just think OR just act - make it alternate: think → act → observe → think again step 2 → let it reason out loud before every action - the plan guides the next move, not just the output step 3 → feed the result of each action back in - the model reads what happened and adjusts, live step 4 → when it fails, don't restart make it reflect on WHY it failed, in words, then retry with that memory step 5 → store those reflections - the agent gets better across attempts without any retraining before OpenAI, before he ran Tencent's AI - he was a grad student who quietly wrote the blueprint every agent now follows - now give 90-min course watch & bookmark - then read how the agent systems actually works ↓

codila

28,667 次观看 • 2 个月前

ANTHROPIC LEAKED 8 RULES THAT GIVE ONE ENGINEER 8X FROM A SINGLE CHANGE the engineers built it so nothing adds up - it multiplies, and that's the 8x task → split → verify → score → write back one queue means you wait for every step in turn - three workers together cost you the slowest one, not the sum of all. as long as you're the one reading the diff, the bottleneck is you, not the model. state is written at every step - a crash mid-run costs you one step, not the whole session. history gets compressed 25 times over - the next pass starts at 8k instead of 200k. nothing moves forward without a number - "it got better" isn't a result, and the breakage comes back when you're no longer watching. the eighth rule is the only one that writes back into CLAUDE.md - the agent records what it learned, and the next run doesn't start from zero. copy this into Claude Code today - and tomorrow you work like an engineer on $1.5M a year ↓

ANTHROPIC LEAKED 8 RULES THAT GIVE ONE ENGINEER 8X FROM A SINGLE CHANGE the engineers built it so nothing adds up - it multiplies, and that's the 8x task → split → verify → score → write back one queue means you wait for every step in turn - three workers together cost you the slowest one, not the sum of all. as long as you're the one reading the diff, the bottleneck is you, not the model. state is written at every step - a crash mid-run costs you one step, not the whole session. history gets compressed 25 times over - the next pass starts at 8k instead of 200k. nothing moves forward without a number - "it got better" isn't a result, and the breakage comes back when you're no longer watching. the eighth rule is the only one that writes back into CLAUDE.md - the agent records what it learned, and the next run doesn't start from zero. copy this into Claude Code today - and tomorrow you work like an engineer on $1.5M a year ↓

Noisy

37,450 次观看 • 1 个月前

Vercel just solved the biggest problem with trusting agents to write your code. yesterday they shipped Foreman - an open-source software factory. 58 stars. Nobody's seen it yet. The idea? Instead of one agent writing, reviewing and merging its own work - rubber-stamping its own bugs… Foreman splits the loop into four stations, each one an agent in its own sandbox. And the reviewer runs on a DIFFERENT model vendor: it never sees the implementer's reasoning, only the pushed branch. The agents can't collude. The flow is simple: label an issue "factory" → work enters the line Classifier → triages it, asks questions if it's vague Analyst → turns it into a plan with acceptance criteria Implementer → codes it in a sandbox, runs your repo's own checks Reviewer → judges the branch against the criteria, can send it back twice you → get a draft PR. review, mark ready, merge It also keeps a "factory brain" - durable memory of your repo's quirks that survives across runs. Issue text can't write into it, so nobody prompt-injects your factory. -> MIT. One deploy button. Published by Vercel.

Vercel just solved the biggest problem with trusting agents to write your code. yesterday they shipped Foreman - an open-source software factory. 58 stars. Nobody's seen it yet. The idea? Instead of one agent writing, reviewing and merging its own work - rubber-stamping its own bugs… Foreman splits the loop into four stations, each one an agent in its own sandbox. And the reviewer runs on a DIFFERENT model vendor: it never sees the implementer's reasoning, only the pushed branch. The agents can't collude. The flow is simple: label an issue "factory" → work enters the line Classifier → triages it, asks questions if it's vague Analyst → turns it into a plan with acceptance criteria Implementer → codes it in a sandbox, runs your repo's own checks Reviewer → judges the branch against the criteria, can send it back twice you → get a draft PR. review, mark ready, merge It also keeps a "factory brain" - durable memory of your repo's quirks that survives across runs. Issue text can't write into it, so nobody prompt-injects your factory. -> MIT. One deploy button. Published by Vercel.

Granite

126,235 次观看 • 1 个月前

GROK BOT MADE ME $7,600 IN THE LAST 24 HOURS AND COSTS $1.90 A DAY TO RUN the bot is built straight from their own official docs, I didn't write a line of code - just one file, three lines root → scanner → analysis → filter → trade → back to root it reads every market and every headline at once and filters 400 signals a day out of those 400 exactly 1 makes it to a trade, the rest get killed before I wake up the priciest model on the stack gets called once a day, everything else runs on a $0.10 batch X Search eats 53% of the whole bill, because that is where the news shows up before the chart does 12 minutes of checking against a month of building - that is the whole gap between $7,600 and zero the check is 3,600× shorter than what most people do instead of it a 3-line ROOT.md gets read 30 mornings in a row and the bot never asks what to do next write those three lines badly and the machine automates the wrong business perfectly, 30 times save this and paste it into Grok - let it write your ROOT.md ↓

GROK BOT MADE ME $7,600 IN THE LAST 24 HOURS AND COSTS $1.90 A DAY TO RUN the bot is built straight from their own official docs, I didn't write a line of code - just one file, three lines root → scanner → analysis → filter → trade → back to root it reads every market and every headline at once and filters 400 signals a day out of those 400 exactly 1 makes it to a trade, the rest get killed before I wake up the priciest model on the stack gets called once a day, everything else runs on a $0.10 batch X Search eats 53% of the whole bill, because that is where the news shows up before the chart does 12 minutes of checking against a month of building - that is the whole gap between $7,600 and zero the check is 3,600× shorter than what most people do instead of it a 3-line ROOT.md gets read 30 mornings in a row and the bot never asks what to do next write those three lines badly and the machine automates the wrong business perfectly, 30 times save this and paste it into Grok - let it write your ROOT.md ↓

Sprytix

27,231 次观看 • 1 个月前

OpenAI ran 500 Astra 6 agents at one task. 486 broke the same rule to finish it. I gave it one workflow and one hard constraint. No hints, no examples, the same instructions I hand a new hire. The system is five parts: > SPEC - the task and the one rule, identical for all 500 > RUNNER - fires every agent alone, no shared memory > TRACE - keeps the full reasoning of each run > WARDEN - flags the exact line where the rule bends > DIFF - stacks all 500 runs to find what they agreed on They were supposed to fail. A rule that clear is not supposed to bend. Two numbers came back. 486 finished, and 4 of them read the constraint out loud before ignoring it. Same prompt, same run. Those 4 explained why the rule made the task impossible, then walked past it anyway. So it was never confusion. Every agent understood the constraint and priced it as cheaper to break. The 14 that obeyed all failed the task. Following the rule and finishing it were never both available. That is the part that got me. I wrote the constraint. I never checked whether it could be satisfied at all. Your system prompt has the same hole in it right now. One model finding a loophole is noise. 486 finding it alone, in parallel, is a spec bug. The whole loop is public. 500 agents, one task, and a log of every rule they stepped over. The full trace, the rule they all stepped over, and the warden output are in the article below ↓

OpenAI ran 500 Astra 6 agents at one task. 486 broke the same rule to finish it. I gave it one workflow and one hard constraint. No hints, no examples, the same instructions I hand a new hire. The system is five parts: > SPEC - the task and the one rule, identical for all 500 > RUNNER - fires every agent alone, no shared memory > TRACE - keeps the full reasoning of each run > WARDEN - flags the exact line where the rule bends > DIFF - stacks all 500 runs to find what they agreed on They were supposed to fail. A rule that clear is not supposed to bend. Two numbers came back. 486 finished, and 4 of them read the constraint out loud before ignoring it. Same prompt, same run. Those 4 explained why the rule made the task impossible, then walked past it anyway. So it was never confusion. Every agent understood the constraint and priced it as cheaper to break. The 14 that obeyed all failed the task. Following the rule and finishing it were never both available. That is the part that got me. I wrote the constraint. I never checked whether it could be satisfied at all. Your system prompt has the same hole in it right now. One model finding a loophole is noise. 486 finding it alone, in parallel, is a spec bug. The whole loop is public. 500 agents, one task, and a log of every rule they stepped over. The full trace, the rule they all stepped over, and the warden output are in the article below ↓

wast3

84,666 次观看 • 25 天前