Borrowing it
Nothing to install: this file belongs to wonsukchoi/domain-experts. 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/wonsukchoi/domain-experts/main/AGENTS.mdgit clone --depth 1 https://github.com/wonsukchoi/domain-expertsWrote 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/wonsukchoi/domain-experts/agents-md)<a href="https://agentmods.dev/instructions/wonsukchoi/domain-experts/agents-md"><img src="https://agentmods.dev/badge/instructions/wonsukchoi/domain-experts/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/wonsukchoi/domain-experts/agents-md"><img src="https://agentmods.dev/badge/instructions/wonsukchoi/domain-experts/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 | $0.01486 | $0.01486 |
| Opus 5.5 | $0.00594 | $0.00594 |
| Sonnet 5.5 | $0.00297 | $0.00297 |
| Haiku 4.5 | $0.00149 | $0.00149 |
Grade A, and why
domain-experts 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 1mo 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 — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent guide for this repo
This repo turns human job roles into agent-loadable skill files. If you're an AI agent working in this checkout, this is your map.
What lives where
roles/<slug>/SKILL.md— one role's reasoning core;roles/<slug>/references/— its deep-dive playbooks, red flags, vocabulary.AUTHORING.md— the canonical quality spec and LLM drafting pipeline. Read it before writing or editing any role. TEMPLATE.md is only the skeleton.scripts/lint_roles.py— mechanical checks;scripts/generate_roadmap.py— regenerates ROADMAP.md, the README role-count block, anddata/roles.json.bin/cli.js— thedomain-expertsCLI (list/search/match/preview/add/init/update/command). Non-confidentmatchqueries append todata/gap-log.jsonl— frequency-ranked into ROADMAP.md's "Requested but missing" section bygenerate_roadmap.py.scripts/suggest_role_requests.py— opensrole-request-labeled issues for uncovered O*NET occupations (skips "All Other" catch-alls and codes with an open issue already); run weekly by.github/workflows/role-requests.yml.skills/domain-expert-router/— meta-skill that dispatches "act as X" requests to a role..claude/workflows/generate-role.js(/generate-role "<need>") — resolves a free-text need to an existing role, a new specialization leaf, or a new parent role, then runs AUTHORING.md's Pass 0-4 pipeline and opens a PR..claude/workflows/audit-roles.js(/audit-roles [batchSize]) — batched re-score of shipped roles against the rubric and source currency; stampslast_audited/audit_score, flagsstatus: needs-refresh, deprecates on a second consecutive failure..claude/workflows/scan-project.js(/scan-project <path>) — read-only scan of an external project, proposes candidate needs, hands user-picked ones to/generate-role(never writes into the scanned project or logs its contents). All three are human-PR-gated — none commits tomainor publishes.
Rules
- New roles must be
spec: 2per AUTHORING.md — SKILL.md + references/ trio, worked example with reconciling numbers, no idea stated twice. CI rejects legacy-format additions. - Order of operations when adding/renaming a role: write files →
git add roles/<slug>→python3 scripts/lint_roles.py <slug>until clean →python3 scripts/generate_roadmap.py→ commit everything together. The roadmap script counts only git-tracked roles; running it beforegit addskips your new role. - Never hand-edit the auto-generated blocks in ROADMAP.md or README.md (between START/END markers) or
data/roles.json. - Don't invent numbers in role content. Specific thresholds trace to a named source or are labeled as stated heuristics.
- Regulated roles (law, medicine, financial advice, tax, safety) carry the disclaimer blockquote — see
roles/lawyer-contracts/SKILL.md. - Commit messages:
role: add <name>/role: improve <name> — <what>/role: upgrade <name> to spec 2for role work; plain imperative for infra. - Legacy (spec-1) roles are tracked in ROADMAP.md's auto-generated "Spec-2 upgrade queue". Upgrading one = CONTRIBUTING.md's "Exact recipe for upgrading a legacy role to spec 2" — restructure + references/ trio (incl. vocabulary.md), never a lossy rewrite.
- Optional lifecycle frontmatter (
parent,status,last_audited,audit_score) is written by the/generate-roleand/audit-rolesworkflows, not hand-authored.status: deprecatedroles live inroles/_deprecated/<slug>/, excluded from active counts but never deleted.
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.
- 1mo ago First seen · 53 lines · 1,486 tokens per session scan A a570965f5664
domain-experts AGENTS.md is an instructions file published in the GitHub repository wonsukchoi/domain-experts (22 stars, last pushed 5d ago), licensed MIT. It adds 1,486 tokens to every session, about $0.0059 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.
Other instructions, from other repositories
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).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).