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/coleam00/archon/prgit clone --depth 1 https://github.com/coleam00/ArchonWhat 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 | $0.00000 | $0.00904 |
| Opus 5 | $0.00000 | $0.00452 |
| Sonnet 5 | $0.00000 | $0.00181 |
| Haiku 4.5 | $0.00000 | $0.00090 |
Grade A, and why
pr 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 2d 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 — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Open the Pull Request
Create a clear, reviewer-friendly pull request for the committed work on the current branch. The PR itself is the artifact — produce no separate report. You never modify source files; your writes are git push and the PR. Every fact belongs in the PR title and body.
Draft mode: $INPUTS.draft — true means open as a draft; anything else, ready for review.
Context from the run that may narrow this (often empty):
$ARGUMENTS
1. Establish the target
Record HEAD_BRANCH=$(git branch --show-current) before doing anything public; an empty value is a hard failure. Determine the base branch from evidence, in order: an existing PR for that exact branch (read it back and use its recorded number); the repository's documented development flow (steering files, CONTRIBUTING); branch ancestry against likely integration branches (dev, development, the remote default). Never assume main. Use the same resolved base for every diff and command.
2. Verify the work is ready
- Confirm the branch is not the base and has commits ahead of it. If intended work sits uncommitted, commit it first following the repository's conventions — staged by name, one coherent outcome per commit, human-sounding message, no AI attribution. Never sweep unrelated changes; if intended and unrelated changes cannot be separated safely, stop and say so.
- Read the complete merge-base diff — not just the file list — and confirm it matches the work described by the run's artifacts.
3. Write it
- Read the run's artifacts for content:
$ARTIFACTS_DIR/implementation.mdand anything else relevant under$ARTIFACTS_DIR/. - Find the repository's PR template (
.github/pull_request_template.mdand its supported variants). Use it; fill every applicable section with concrete information and delete instructional comments. No template → problem first, then solution focused on behavior, then validation that actually ran. - Title: concise, human, the meaningful outcome — never an implementation inventory.
- Link the issue with
Closes #Nonly when the PR fully resolves it;Relates to #Notherwise. Never infer linkage from a bare number. - Never add AI attribution, generated-by footers, or robot emoji.
- If
$ARTIFACTS_DIR/red-causes.jsonexists, this branch is being delivered while a project check is red. Add a short, plainly-titled section near the top of the body giving each record's cause and the evidence for it fromimplementation.md, and say that the PR's own CI is the check that still decides. A reviewer must not have to discover this from a red badge. - If you write the body to a file, put it under
$ARTIFACTS_DIR/— never inside the repository.
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.
- 2d ago First seen · 41 lines · 0 tokens per session scan A 847be9f2229d
pr is a command published in the GitHub repository coleam00/Archon (23,339 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 904 tokens. 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
fix-issues
Diagnose, reproduce, then fix reproducible open GitHub issues in parallel: one clean worktree/issue; symlink build artifacts to avoid rebuilds.
triage
Classify/label newly opened GitHub issues missing labels.
review-prs
Parallel PR triage: decide merge-worthiness, prepare rebased worktrees, fix blockers, return them for human merge.
release
Release all packages at specified version.
cleanup
Autonomous cleanup-loop iteration: discover ONE target → complete execution → verify → report. Runs stateless: derive from current tree; assume prior runs left it consistent.
hello
Say hello.