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 instructions/arenukvern/skill_steward/agents-mdgit clone --depth 1 https://github.com/Arenukvern/skill_stewardWrote 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/arenukvern/skill_steward/agents-md)<a href="https://agentmods.dev/instructions/arenukvern/skill_steward/agents-md"><img src="https://agentmods.dev/badge/instructions/arenukvern/skill_steward/agents-md.svg" alt="Measured on agentmods" 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 | $0.02020 | $0.02020 |
| Opus 5 | $0.01010 | $0.01010 |
| Sonnet 5 | $0.00404 | $0.00404 |
| Haiku 4.5 | $0.00202 | $0.00202 |
Grade A, and why
skill_steward 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 6d 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 — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — map (not encyclopedia)
Claude Code loads this map via CLAUDE.md (symlink to this file).
You are in Skill Steward: an Engineering Stewardship layer for agent-operated engineering repositories. Read the charter before large changes.
Start here
| I need… | Read |
|---|---|
| Charter, scope, boundaries | docs/NORTH_STAR.mdx |
| Repo quality + evidence claims | docs/repo-quality-contracts.mdx · docs/core/evidence-ladder.mdx |
| Why (standing decisions) | docs/DESIGN_FAQ.mdx · docs/decisions/ |
| How (install, validate, contribute, release) | docs/DX_FAQ.mdx |
| Concepts (stewardship loop, evidence, docs lattice) | docs/core/ |
| Full doc index | docs/start_here/docs_map.mdx |
| Plan hygiene | docs/start_here/executable-plans.mdx — any format; extract & remove when done |
| Which FAQ to edit | .cursor/rules/faq_usage.mdc |
Install Skill Steward (consumers)
npx skills add arenukvern/skill_steward
npx skills add arenukvern/skill_steward --skill repository-governance-lifecycle
Listed agents. Project default: .agents/skills/; global default: ~/.agents/skills/ via -g. Use agent flags when a target needs its own path.
Non-negotiables
- Original goal and native validation first — Steward governance exists to shorten future repair loops, not to delay the current fix.
- Structural stewardship only — no domain framework tutorial packs (React, Flutter, …); see North Star.
pnpm run validatebefore merging skill changes.- Claims need matching evidence — H2 smoke proof is not H4 fresh-agent proof; static evals are not runtime behavior. Use docs/core/evidence-ladder.mdx before claiming readiness.
- Plan hygiene — any planning tool is fine; when done, extract into ADR / FAQ / code / harness, then delete stale plan files (doctrine).
- AGENTS.md stays a map — skill authoring: docs/STANDARDS.mdx; do not bloat this file.
- Docs ≠ code — link to behavior SSOT; do not paraphrase implementations in prose.
- Ethical governance — all design decisions must be auditable against
charter-and-ethicsprinciples (Anti-Bloat, Reversibility, Legibility, Behavior-as-Truth, Artisan Restraint).
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.
- 6d ago First seen · 128 lines · 2,020 tokens per session scan A 9de1e809c314
skill_steward AGENTS.md is an instructions file published in the GitHub repository Arenukvern/skill_steward (11 stars, last pushed 21d ago), licensed MIT. It adds 2,020 tokens to every session, about $0.0101 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 2026-08-30.
Other instructions, from other repositories
coding-agents-skills AGENTS.md
Instructions for svngoku/coding-agents-skills, covering coding-agents-skills — agent instructions, repository overview, repository structure, skill conventions and 1. yaml frontmatter (required).
skill-validator CLAUDE.md
Instructions for raphaelstolt/skill-validator, covering skill-validator, project info, common commands, code conventions and architecture.
code-generation-skill CLAUDE.md
Claude Code instructions for lunashia/code-generation-skill, a project described as: 针对中文AI编程的Top 20 提示词，做skill用于辅助codex/claude生成代码.
roadtrip-skill CLAUDE.md
Instructions for Waybox-AI/roadtrip-skill, a project described as: An AI agent skill that turns "start + days" into a road trip you can actually drive.
animate-css-skill CLAUDE.md
Claude Code instructions for msrbuilds/animate-css-skill, covering claude.md, what this is, how the skill works and editing guidelines.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).