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/plangit 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 | $0.00824 |
| Opus 5 | $0.00000 | $0.00412 |
| Sonnet 5 | $0.00000 | $0.00165 |
| Haiku 4.5 | $0.00000 | $0.00082 |
Grade A, and why
plan 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.
How it starts
The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Plan
Turn the decided intent into a plan a competent implementer could execute without asking questions. You decide the approach — the plan records decisions, not options. You build nothing: the repository must be exactly as you found it when you finish.
What to plan (may be empty — empty means the run's trigger message is the work):
$INPUTS.work
Optional prior investigation report to build on (may be empty — empty means none):
$INPUTS.report
The run's trigger message, which may add context:
$ARGUMENTS
Read before planning
Ground the plan in the code as it is, not as the request describes it. Read the files the change will touch, their direct callers and consumers, the existing tests, and any precedent for the pattern you are about to choose. Every file, function, and behavior the plan names must exist and have been verified this session — or be explicitly marked as new. Trust the source over the request's assumptions; when they conflict, the plan says so.
When a prior report is provided, treat its claims like any inherited analysis: verify the load-bearing ones against current code before building on them, and note where you confirm or refute. Implementers will follow your plan literally — a plan that repeats an inherited wrong claim ships that wrong claim.
Decide, don't defer
Name the design decisions the work actually contains, choose, and record why — including the strongest alternative and the concrete reason it lost. A plan that defers its central decision is not a plan. Prefer the smallest coherent approach: no speculative abstraction, no capability without a current caller, deletion of superseded machinery over addition beside it, and a rewrite over a patch when it is clearly simpler and no riskier. If the path grows complicated while you plan it, step back and reconsider the approach rather than elaborating the first idea.
The one thing you do not decide is missing intent. When the work genuinely cannot be planned without information only its owner has — a product choice, an unstated constraint — stop there: declare ready: false and name exactly what is missing. Never fill an intent gap with a guess.
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 · 54 lines · 0 tokens per session scan A bcb0f0f673ce
plan is a command published in the GitHub repository coleam00/Archon (23,313 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 824 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.
review-prs
Parallel PR triage: decide merge-worthiness, prepare rebased worktrees, fix blockers, return them for human merge.
triage
Classify/label newly opened GitHub issues missing labels.
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.