agy
Skill1 starsChanged 25 days agoMITView on GitHub
Delegate bulk work and second opinions to Antigravity (agy).
10 other projects use a version of this.
Adopted by
Most recently taken up by
- fmind/dottheir own version, first seen Nov 8, 2026
- fmind/dotfilestheir own version, first seen Nov 8, 2026
- letitbk/claude-academic-setuptheir own version, first seen Nov 1, 2026
- mustafarslan/skillstheir own version, first seen Nov 1, 2026
- darshitpp/x-agenttheir own version, first seen Oct 13, 2026
- laurentvv/agents-kittheir own version, first seen Oct 11, 2026
Used with
Projects with it also use codex.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
engtySkillcodex
Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex for code analysis, refactoring, or automated editing
Also in 4 projects that use this
Other versions10 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 10
- letitbk/claude-academic-setup★ 48same name
- fmind/dot★ 13same name
- fmind/dotfiles★ 13same name
- phuongduyphan/cc-tmux-agents★ 8same name
- darshitpp/x-agent★ 2same name
- mustafarslan/skills★ 2same name
- shauryaaojha/claude-code-delegate★ 2same name
- ssv445/claude-skills★ 2same name
- dT-Tb-labs/babel-orchestration★ 1same name
- laurentvv/agents-kit★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 8 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 55
Never trust a self-reported "done".
- Distinctive: line 63
Transcripts: treat flagged-unclear passages as unverified.
“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
agy
Hand work to Google's Antigravity CLI (agy, running Gemini) and keep the
judgment here. You stay the conductor: you own requirements, verification, and
the final call. agy is the executor.
Read references/invariants.md before your first
agy call in a session. It carries the flag rules that a plausible-looking
invocation gets silently wrong — the failure mode is a confident wrong answer,
not an error, so guessing the syntax is the expensive mistake here.
When to use
Delegate when the offloaded volume clearly exceeds the cost of specifying the task, the round trip, and verifying the result. Bulk, repetitive, or parallel work qualifies. A two-line edit does not — doing it yourself is cheaper.
Also use it for things this agent cannot do at all: hearing audio, watching video, and getting a genuinely independent second opinion from a different model family.
Operations
| operation | when | how |
| setup | first run, or agy misbehaving | references/setup.md |
| review | before commit or merge | references/review.md |
| delegate | bulk build, scaffolding, tests | references/delegate.md |
| research | multi-source web questions | references/research.md |
| media | audio, video, images | references/media.md |
Two rules that decide whether this pays off
Ask for a digest, never raw output. Close every read/analysis prompt with an explicit contract — "Output ONLY 5–8 bullets, no file contents." This is the whole cost saving.
Never re-read what agy already handled. Ingest the digest and move on. Re-reading the files puts back exactly the cost you just removed.
Verification gates
Never trust a self-reported "done".
- Writes: confirm with
rtk proxy git status(orgit statusif RTK is not installed) that files actually changed. Run write tasks on a dedicated branch. - Review findings: corroborate each one against the real code before acting. Cross-model agreement is a strong signal; disagreement means look closer.
- Research citations: open the URLs. agy can present parametric knowledge as a sourced fact.
- Transcripts: treat flagged-unclear passages as unverified. Read the timestamp slice before relying on any load-bearing figure or quote.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.