Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/marcosnahuel/antigravity-plugin-cc/graphgit clone --depth 1 https://github.com/MarcosNahuel/antigravity-plugin-ccWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/marcosnahuel/antigravity-plugin-cc/graph)<a href="https://agentmods.dev/commands/marcosnahuel/antigravity-plugin-cc/graph"><img src="https://agentmods.dev/badge/commands/marcosnahuel/antigravity-plugin-cc/graph.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00106 | $0.01434 |
| Opus 5 | $0.00053 | $0.00717 |
| Sonnet 5 | $0.00021 | $0.00287 |
| Haiku 4.5 | $0.00011 | $0.00143 |
Grade A, and why
graph scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Turn a folder into a queryable knowledge graph using Graphify.
Since Graphify 0.9 the code graph is deterministic and local — tree-sitter ASTs, no LLM, no API
key, no tokens on anyone's budget. The one step that wants a language model is naming the communities,
and this plugin hands that to Gemini via agy (one call), so the host assistant still spends
nothing. Claude then reads the small graph.json / GRAPH_REPORT.md to reason — no re-reading the
raw files.
Raw user request: $ARGUMENTS
Phase 0 — Ensure the stack (ONE Bash call)
Parse $ARGUMENTS: the leading existing-directory token is the folder. Then:
python "${CLAUDE_PLUGIN_ROOT:-$PWD}/plugins/antigravity/scripts/graphify_install.py"
- Last line
READY→ continue with everything below. READY_NO_AGY→ continue, but skip Phase 2 (communities keep hub-derived names). Mention that installing + signing intoagy(https://antigravity.google) gets them named.MISSING/ non-zero exit → relay the printed reason and stop.- Also capture the
GRAPHIFY_CMD=line — it is how to invoke Graphify on this machine (the console script, or"<python>" -m graphifywhen the Scripts dir is not on PATH). Use it verbatim below as$GFY.
Phase 1 — Build the graph (local, free)
eval "$(python "${CLAUDE_PLUGIN_ROOT:-$PWD}/plugins/antigravity/scripts/graphify_outdir.py" "<FOLDER>")"
$GFY extract "<FOLDER>" --code-only
graphify_outdir.pyis the Windows MAX_PATH guard: past ~140 characters of project path the AST cache filename crosses Windows' 260-char ceiling, and Graphify then prints "AST extraction failed" + "graph is empty" and still exits 0. For deep projects it emits anexport GRAPHIFY_OUT=...line (relocating the outputs, and telling you where on stderr); for short paths and non-Windows it prints nothing, so theevalis a no-op. Keep the same environment for every later phase — the sidecars live wherever this pointed.- Note the
found N code, M docs, P papers, Q imagesline — you needM+P+Qin Phase 4. - Honest gate: if the output says
graph is emptyor0 nodes, the run FAILED. Say so and stop; do not present an empty graph as a result. Exit code 0 does not mean success here.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 103 lines · 106 tokens per session scan A eb611f2596ac
graph is a command published in the GitHub repository MarcosNahuel/antigravity-plugin-cc (26 stars, last pushed 18d ago), licensed MIT. It adds 106 tokens to every session and 1,434 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
gemini-swarm
Swarm a task across parallel Gemini (agy) sessions with automatic decomposition and model routing.
gemini-imagegen
Generate or edit an image via a Gemini (agy) session — Gemini's generateimage tool does what Claude cannot.
agy-debug
Send an error/stack trace plus the relevant file(s) to Antigravity (agy) for root-cause analysis.
agy-review
Send the current change set to Antigravity (agy) for a senior-reviewer code review.
agy-panel
Ask Antigravity (agy) the same question across several distinct models and compare their answers side by side.
agy
Consult Google's Antigravity CLI (agy) and relay its answer.