Borrowing it
Nothing to install: this file belongs to cesarferreira/stax. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/cesarferreira/stax/main/AGENTS.mdgit clone --depth 1 https://github.com/cesarferreira/staxWrote 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/instructions/cesarferreira/stax/agents-md)<a href="https://agentmods.dev/instructions/cesarferreira/stax/agents-md"><img src="https://agentmods.dev/badge/instructions/cesarferreira/stax/agents-md/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/instructions/cesarferreira/stax/agents-md"><img src="https://agentmods.dev/badge/instructions/cesarferreira/stax/agents-md.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 | $0.02562 | $0.02562 |
| Opus 5.5 | $0.01025 | $0.01025 |
| Sonnet 5.5 | $0.00512 | $0.00512 |
| Haiku 4.5 | $0.00256 | $0.00256 |
Grade A, and why
stax AGENTS.md 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 21d 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 — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
stax: Rust CLI for stacked Git branches and PRs. Branch metadata lives in git refs (refs/branch-metadata/<branch>, trunk in refs/stax/trunk), compatible with freephite. Everything else is discoverable from src/; this file only covers what isn't.
Harness: stax-dev
Goal: Accelerate stax feature development and bug fixing via a git-native worktree → plan → implement → review → verify → (optional) benchmark → commit → draft-PR pipeline with a bounded repair loop.
Trigger: For any code change to the stax CLI (new commands, flags, bugfixes, refactors, behavior changes), run the /stax-dev prompt. It orchestrates the project-local stax-planner, stax-implementer, stax-reviewer, stax-verifier, stax-benchmarker, and stax-release-manager definitions under .claude/agents/. The former .pi/agents/ definitions were intentionally moved to personal dotfiles and are not the project harness. Direct usage/architecture questions don't need the pipeline. Each task runs in its own worktree on a stacked branch; commit + current-branch-only draft PR submission (stax branch submit --draft --ai --yes) is automatic when the applicable gate is green; merging to main is never automatic and promoting a draft to ready-for-review is HITL-gated.
Change History:
| Date | Change | Target | Reason |
|---|---|---|---|
| 2026-07-25 | Initial harness build (5 agents, 5 skills, /stax-dev orchestrator) |
All | New harness |
| 2026-07-25 | Git-native evolution: added benchmarker (optional perf gate) + stax-benchmark skill; worktree-per-task + stacked branch; auto commit-on-green + auto draft PR; hard "never merge to main" rule |
benchmarker.md, stax-benchmark/, release-manager.md, codex.md, stax-release/, stax-dev.md | Adopt git-native workflow constraints; keep verifier as mandatory correctness gate |
| 2026-07-28 | Added fallbackModels: claude-bridge/claude-sonnet-4-6 to the four cursor/composer-2.5 agents so a provider outage (quota/auth/timeout) auto-routes instead of stalling the run; codex also falls back to openai-codex/gpt-5.6-terra. Extended benchmarker skip rule to cover pure delegation/deletion diffs with no new local hot-path work |
codex.md, verifier.md, benchmarker.md, release-manager.md | Cursor was down + OpenAI-Codex usage-limited mid-run (stax-dev-03), forcing manual model overrides; benchmarker also burned time trying to bench a no-local-compute change |
| 2026-08-26 | Split verification into a fast scoped draft gate and a full CI/ready-for-review gate | AGENTS.md, verifier.md, stax-verify/, stax-dev.md | Avoid repeating the process-heavy full integration suite in every isolated worktree while retaining full validation before review or merge |
| 2026-08-27 | Reconciled the project-local harness with its documented git-native pipeline: added independent review, optional benchmark, bounded repair artifacts, and release ownership with current-branch-only draft submission | .claude/agents/, .claude/skills/stax-dev/, skills.md, AGENTS.md |
Restore the missing commit/draft-PR path and keep full-stack submit, undraft, and merge outside automation |
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.
- 21d ago Changed · +54 lines · +1,677 tokens per session bd2a11807207
- 1mo ago Changed · -107 lines · -1,896 tokens per session fe91b5ebde82
- 1mo ago Changed · +80 lines · +953 tokens per session 4413bb6aaf9d
- 2mo ago First seen · 69 lines · 1,828 tokens per session scan A a884381c1d76
stax AGENTS.md is an instructions file published in the GitHub repository cesarferreira/stax (133 stars, last pushed 2d ago), licensed MIT. It adds 2,562 tokens to every session, about $0.0102 per session on Opus 5.5. 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.
stax AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| stax AGENTS.mdcesarferreira/stax | Instructions file | 133 | 2d ago | MIT | A | 2,562 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 274k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |