Skip navigation
Antigravity Agent handles skill calling two ways: implicit automation that anticipates your needs…
0:00 / 0:52

Antigravity Agent handles skill calling two ways: implicit automation that anticipates your needs…

Google Antigravity

@antigravity222K followers

Visit on X
1,509 likesDownloadOriginal post

134,891 views

Antigravity Agent handles skill calling two ways: implicit automation that anticipates your needs, and explicit /skillname commands for when you want total control. It’s the perfect transition between AI automation and human intent. See how it works. 👇

34 Comments

Comments from the original post on X (Twitter)

  • Mahendra Kumawat's profile picture

    @Mk__0168

    My antigravity 😭

    26 likes

  • userisme's profile picture

    @quietlogic22

    still can't use it... for weeks now. despite paying for "AI Ultra" please fix this, it's been long enough.

    9 likes

  • Silvan Soeters's profile picture

    @silvansoeters

    sweett agent manager gets "stuck in a loop" often btw

    7 likes

  • noadrev's profile picture

    @noadrev

    @antigravity The upgrade to the latest 1.20.5 broke the slash command functionality and rules and workflows no longer show up in the Customisations UI, even when using the skills standards. Downgrading to 1.19.6 restores lost functionality. Hopefully fixed in an upcoming release

    7 likes

  • Builds After 5 🦉's profile picture

    @buildsafter5

    claude code has the same /skill pattern. the real differentiator isn't the syntax - it's how good the implicit detection is. how often does it guess wrong on ambiguous prompts?

    4 likes

  • Jose.ink 🚀☀️🌴's profile picture

    @Jojocryptoquest

    @antigravity hey guys could you please fix the bug that doesn’t let us see our most recent conversations? Thank you for your attention to this matter

    4 likes

  • Bruno's profile picture

    @bfms1313

    What i get everytime i submit, any prompt. Trajectory ID: 1749d14b-7a74-498e-80f8-c91c3f485c24 Error: agent executor error: could not convert a single message before hitting truncation So.. Antigravity is dead..

    3 likes

  • Melih Toksari's profile picture

    @melihtoksari

    Skills are game changing agentic behavior, taking superhuman serum!

    3 likes

  • polihedge's profile picture

    @polihedge

    Antigravity is sometimes aware and sometimes not aware of the Agents and Skills standards. It’s totally inconsistent and programmed poorly. Amateurish.

    3 likes

  • Michele Guizzardi's profile picture

    @micheleguizzard

    i have antigravity and gemini pro. I got a 5 days lock instead of 5h Your plan's baseline quota will refresh on 16/03/2026, 13:27:04. @antigravity

    3 likes

  • Nova's profile picture

    @Nova_Nightfall

    I encountered a five-hour limit six hours ago, and it just reset, now it‘s at 20% and refreshed weekly 🙄

    2 likes

  • nonehere's profile picture

    @testerhahaha

    Guys can u please get back Claude Opus 4.6 Token Window Capacity to it's default which is 1M token !? Because now it is only 200k (200000)...?

    1 like

  • Praneth's profile picture

    @praneth02

    @antigravity Fix your rate limits. Since January, many AI Pro users are seeing days of wait time instead of the promised 4–5 hour refresh, even without using the model. Paid users shouldn’t be treated like the free tier. Please fix this ASAP.

    1 like

  • ソロプレナー日誌's profile picture

    @qd4dky

    When a SKILL is invoked, it is always named “SKILL.md”. In contrast, workflows are named “workflow_name.md”. It would be much more user-friendly if it were possible to clearly indicate which SKILL is being invoked.

    1 like

  • いわ(iwa_30)'s profile picture

    @iwa_30

    Connecting to language server... It won't go on like this...

    1 like

  • LIPPP's profile picture

    @LIPPPX0

    I deliberately let it, so that the tokens can be used up. look at what he analyzes, he's just hallucinating, global giant technology but stupid AI thinking features and a thousand hallucinations. the loss of subscribing only makes the tokens run out because of the stupid AI model

    1 like

  • Partha A's profile picture

    @sparthan18

    latest update 1.20.5 is very much unable to use, only response it gives is Error Unknown: Agent execution terminated due to error. @OfficialLoganK

    1 like

  • FindLogan's profile picture

    @FindLogan

    This is a massive win towards the right direction. For people using skills elsewhere, having the ability for calling skills via the chat is SUPER helpful.

    1 like

  • Mykyta Pavlenko's profile picture

    @mktpavlenko

    strong pattern. would be interesting to see the discoverability split, how many people stay in the implicit flow vs still reaching for slash commands

    1 like

  • Lukas L's profile picture

    @LukasL_Hater

    Fix your antigravity quata!! or refund!!