Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/metagit-ai/metagit-clinpx agentmods add skills/metagit-ai/metagit-cli/metagit-graph-maintainWrote 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/skills/metagit-ai/metagit-cli/metagit-graph-maintain)<a href="https://agentmods.dev/skills/metagit-ai/metagit-cli/metagit-graph-maintain"><img src="https://agentmods.dev/badge/skills/metagit-ai/metagit-cli/metagit-graph-maintain/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/metagit-ai/metagit-cli/metagit-graph-maintain"><img src="https://agentmods.dev/badge/skills/metagit-ai/metagit-cli/metagit-graph-maintain.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00049 | $0.01132 |
| Opus 5 | $0.00024 | $0.00566 |
| Sonnet 5 | $0.00010 | $0.00226 |
| Haiku 4.5 | $0.00005 | $0.00113 |
Grade A, and why
metagit-graph-maintain 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 9d 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 — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Workspace graph maintenance
Use this skill to discover and keep durable cross-repo edges in .metagit.yml graph.relationships, then sync them into GitNexus.
When to use
| Phase | Trigger |
|---|---|
| Discover | New umbrella workspace, empty graph.relationships, or first graph authoring pass |
| Maintain | After adding repos, promoting inferred deps, or refreshing GitNexus overlay |
First-time discovery (report only)
Do not apply until the operator approves the discovery report.
export METAGIT_AGENT_MODE=true
metagit prompt workspace -c .metagit.yml -k graph-discover --text-only
metagit workspace list -c .metagit.yml --json
metagit config graph suggest -c .metagit.yml --json --include-declared --min-confidence all
Deliver a Graph Discovery Report:
- inferred edges grouped by confidence (
high/medium/low) - gap pairs with no machine path (need operator interview)
proposed_manual_edges[]from interview answersoperations_previewJSON — not applied
After sign-off, continue with the maintenance workflow below (or metagit prompt workspace -k graph-maintain).
Maintenance workflow
CLI flags: -c on config graph suggest|export is the manifest path (trailing leaf -c is valid). --workspace-root is the checkout root used to scan inferred deps. Global metagit -c configures appconfig, not .metagit.yml.
1. Bootstrap context
export METAGIT_AGENT_MODE=true
metagit prompt workspace -c .metagit.yml -k graph-maintain --text-only
metagit workspace list -c .metagit.yml --json
2. Suggest candidates
metagit config graph suggest -c .metagit.yml --json
metagit config graph suggest -c .metagit.yml --min-confidence high --json
MCP: metagit_suggest_graph_relationships
Review candidates[] for confidence, evidence, and source_edge_type. Skip low-confidence edges unless the operator approves.
Also review stale_manual[] — active manual relationships with no supporting inferred edge under the current scan, matched on endpoints so a differing relationship type still counts as support (report-only; do not edit/remove without operator confirmation). Use --verbose when candidates are empty to confirm scan roots and ignore prune counts.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago First seen · 136 lines · 49 tokens per session scan A 8a4628f7e73b
metagit-graph-maintain is a skill published in the GitHub repository metagit-ai/metagit-cli (3 stars, last pushed today), licensed MIT. It adds 49 tokens to every session and 1,132 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
clawrium-release-announcements
Daily — draft a release blog post, open a PR, iterate on comments until merged.
blog-author
Watch ric03uec/clawrium release tags; draft a short blog post per user-visible feature as a PR against blog/.
release-watcher
Watch upstream Claw releases and clawrium discussions; surface top 3 feature candidates to Devashish via Discord DM for approve/skip.
perseus
Use when you need a bounded, local Perseus context render before an assistant reads project state. Perseus resolves selected workspace inputs such as git, services, sessions, and task notes into markdown. Use for deterministic session starts, workspace audits, and explicit context handoffs.
pypi-release
This skill should be used when releasing tunacode-cli to PyPI. It keeps the existing local release checks, then hands the actual PyPI upload to a GitHub Actions workflow that uses the repository's PYPIAPITOKEN secret.
memory-commit
Use when the user explicitly says "remember this", "save this", "ghi nho", "luu lai", "save for next time", or otherwise asks to persist the immediately preceding context. Captures with the appropriate contexttype (decision, preference, fact, skill, task, conversation) so future sessions can retrieve it accurately.