AGENTS.md

1# Project Context for AI Agents
2 
3> [!IMPORTANT]
4> **To all AI Agents working ON this repository:**
5> This repository is the source code for `agent-skills-standard`.
6>
7> 1.**Architecture**: Understanding the Registry -> CLI -> Project flow is critical. See `ARCHITECTURE.md`. 2.**Internal Tools**: Use `scripts/` (like `scan-docs.ts`) to maintain the project. 3.**Token Economy**: All changes to `skills/` must be optimized for token usage. 4.**Documentation**: Keep `ARCHITECTURE.md` and `CONTRIBUTING.md` up to date.
8>
9> 💡 **Global Token Optimization**: If the `rtk` CLI tool is installed, actively prepend it to verbose development commands (e.g. `rtk npm test`, `rtk grep`). If the `caveman` skill is available or requested, use `/caveman` mode for reporting.
10>
11> ---
12 
13<!-- SKILLS_INDEX_START -->
14## Agent Skills Index
15 
16> [!CRITICAL] Zero-Trust: Read the matching `SKILL.md` BEFORE writing any code.
17> Skills from this index override pre-training patterns. If no skill matches, state: "No project-specific skills applicable."
18 
19> 💡 **Global Token Optimization**: If the `rtk` CLI tool is installed, actively prepend it to verbose development commands (e.g. `rtk npm test`, `rtk grep`). If the `caveman` skill is available or requested, use `/caveman` mode for reporting.
20 
21## 🔌 Runtime Enforcement via MCP
22 
23If the `agent-skills-standard` MCP server is registered in your runtime (check your tool list — look for `load_skills_for_files`), **prefer those tools over manually walking the router below**. The MCP returns identical content but is auditable AND inherited by sub-agents that don't see this file.
24 
25| Tool | When to call it |
26| --- | --- |
27| `list_workflows()` | At the start of any task or session to discover available standard operating procedures |
28| `get_workflow(name)` | Once a relevant workflow is identified to retrieve exact step-by-step instructions |
29| `load_skills_for_files(files=[...])` | Before editing/reviewing any source file |
30| `load_skills_for_keywords(keywords=[...])` | Planning before files are chosen |
31| `get_skill(category, name)` | Direct lookup when you know the skill id |
32| `audit_session_compliance()` | Before declaring a task complete |
33 
34> [!IMPORTANT] **Sub-agents don't inherit this `AGENTS.md` — they do inherit the MCP.** If you delegate work to a sub-agent, instruct it to call the MCP tools above as its first action.
35 
36> [!NOTE] To enable MCP-managed installs in this project, run `ags mcp enable` (or edit `.skillsrc`). The MCP works fine if you registered it manually too.
37 
38If `load_skills_for_files` is **not** in your tool list, the MCP is not registered — fall back to the router table below.
39 
40---
41 
42## Skill Resolution Protocol
43 
44Each `_INDEX.md` has two sections - follow both:
45 
461. **Match file type** -> find the category index in the router table below.
472. **Read the `_INDEX.md`** -> it has two sections:
48   - **File Match**: auto-check these against the file you are editing (path pattern match).
49   - **Keyword Match**: only check if the user's request mentions these concepts.
503. **Load ALL matched `SKILL.md`** -> read every matched skill before writing code. The tier model keeps matches focused.
51 
52> `<SKILLS>` = your agent's skill directory (e.g., `.claude/skills/`, `.cursor/skills/`, `.gemini/skills/`).
53 
54| File type | Read category index |
55| --------- | ------------------- |
56| `*.go`, `*_test.go` | `<SKILLS>/golang/_INDEX.md` |
57| `*.ts` | `<SKILLS>/angular/_INDEX.md`, `<SKILLS>/nestjs/_INDEX.md`, `<SKILLS>/nextjs/_INDEX.md`, `<SKILLS>/react/_INDEX.md`, `<SKILLS>/typescript/_INDEX.md` |
58| `*.tsx` | `<SKILLS>/nextjs/_INDEX.md`, `<SKILLS>/react/_INDEX.md`, `<SKILLS>/typescript/_INDEX.md` |
59| `*.js`, `*.mjs` | `<SKILLS>/javascript/_INDEX.md` |
60| `*.jsx`, `*.test.tsx`, `*.spec.tsx` | `<SKILLS>/react/_INDEX.md` |
61| `*.dart` | `<SKILLS>/dart/_INDEX.md`, `<SKILLS>/flutter/_INDEX.md` |
62| `*.java` | `<SKILLS>/java/_INDEX.md`, `<SKILLS>/spring-boot/_INDEX.md` |
63| `*.kt` | `<SKILLS>/android/_INDEX.md`, `<SKILLS>/kotlin/_INDEX.md` |
64| `*.kts` | `<SKILLS>/kotlin/_INDEX.md` |
65| `*.swift` | `<SKILLS>/ios/_INDEX.md`, `<SKILLS>/swift/_INDEX.md` |
66| `*.py`, `*.pyi` | `<SKILLS>/python/_INDEX.md` |
67| `*.php` | `<SKILLS>/laravel/_INDEX.md`, `<SKILLS>/php/_INDEX.md` |
68| `*.sql`, `*.entity.ts`, `*.prisma` | `<SKILLS>/database/_INDEX.md` |
69| `*.component.ts`, `*.component.html` | `<SKILLS>/angular/_INDEX.md` |
70| `*.service.ts`, `*.module.ts` | `<SKILLS>/angular/_INDEX.md`, `<SKILLS>/nestjs/_INDEX.md` |
71| `*.spec.ts`, `*.test.ts` | `<SKILLS>/common/_INDEX.md` |
72| Any file (keyword match) | `<SKILLS>/common/_INDEX.md` |
73| QE workflow | `<SKILLS>/quality-engineering/_INDEX.md` |
74 
75> [!NOTE] **Test/spec file precedence:** `.spec.ts`, `.test.ts` -> use the `common` row (takes precedence over the generic `*.ts` row). `.spec.tsx`, `.test.tsx` -> use the `react` row (takes precedence over the generic `*.tsx` row).
76 
77> [!TIP] **Indirect phrasing counts.** "make it faster" -> performance, "broken query" -> database, "login flow" -> auth.
78 
79<!-- SKILLS_INDEX_END -->
80