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/sciencemj/graph-engineering/graph-rungit clone --depth 1 https://github.com/sciencemj/graph-engineeringWhat 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 | $0.00010 | $0.00241 |
| Opus 5 | $0.00005 | $0.00120 |
| Sonnet 5 | $0.00002 | $0.00048 |
| Haiku 4.5 | $0.00001 | $0.00024 |
Grade A, and why
graph-run 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 yesterday.
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.
What it actually says
Follow the pump loop and Never sections of the graph-engineering skill exactly.
Target: $ARGUMENTS
- Create the run with
run_start. Nothing executes at this point. - Call
run_advancerepeatedly. Handle each response kind as the skill's pump loop says:executed→ reportstepsbriefly and continue. If a step hasstatus:"running", that command is still going — callrun_advanceagain right away to collect it.ai_task→ readcontext, carry outinstructionand only that, thenrun_submit.waiting_user→ tell the user and stop. Do not poll.finished→ final summary.
Remember: you are a pump, not a scheduler. The server decides which node runs.
When an ai_task fails, report the failure as-is — retries are declared in the graph and enforced by
the server.
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.
- yesterday First seen · 21 lines · 10 tokens per session scan A 95ed761430fc
graph-run is a command published in the GitHub repository sciencemj/graph-engineering (0 stars, last pushed 1mo ago), licensed MIT. It adds 10 tokens to every session and 241 once invoked, about $0.0001 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 commands, from other repositories
upgrade-webkit
Upgrade Bun's WebKit fork to the latest upstream version of WebKit.
dedupe
Find duplicate GitHub issues.
autoprove
Autonomous multi-cycle theorem proving with hard stop rules.
prove
Guided cycle-by-cycle theorem proving with explicit checkpoints.
draft
Draft Lean declaration skeletons from informal claims.
rp-build-cli
Build with rp-cli context builder → chat → implement.