oliver-kriska/claude-elixir-phoenix / .claude/skills/docs-check
docs-check
VerifiedCONTRIBUTOR TOOL - Validate plugin against latest Claude Code documentation. Catches breaking changes, deprecations, discovers new features. Run before releases or periodically. NOT part of the distributed plugin.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.2 KB |
| references/doc-pages.md | markdown | 2.2 KB |
| references/validation-rules.md | markdown | 5.7 KB |
references/validation-rules.md
View on GitHub# Validation Rules
Per-component checklists for validation workers.
Each section is passed ONLY to the subagent responsible for that component type.
## Agent Validation Rules
### Frontmatter Fields
Check each agent `.md` YAML frontmatter against `sub-agents.md` docs.
**Required fields:**
- `name` — lowercase, hyphens only, no spaces
- `description` — must include when to use/delegate guidance
**Optional fields (check valid values if present):**
| Field | Valid Values | Notes |
|-------|-------------|-------|
| `model` | `sonnet`, `opus`, `haiku`, `inherit` | Default: inherit |
| `permissionMode` | `default`, `acceptEdits`, `dontAsk`, `bypassPermissions`, `plan` | |
| `tools` | `Read`, `Write`, `Edit`, `Bash`, `Grep`, `Glob`, `Agent`, `WebFetch`, `WebSearch`, `NotebookEdit`, `Skill`, `AskUserQuestion`, `TaskCreate`, `TaskUpdate`, `TaskOutput`, `KillShell`, `MCPSearch`, `ExitPlanMode` | Check docs for new tools |
| `disallowedTools` | Same tool names as `tools` | |
| `maxTurns` | Positive integer | |
| `skills` | List of skill names | Verify referenced skills exist |
| `mcpServers` | Object or list | |
| `hooks` | Object | Per-agent lifecycle hooks |
| `memory` | `user`, `project`, `local` | Auto-enables Read/Write/Edit |
| `background` | `true`, `false` | Always run as background task |
| `isolation` | `worktree` | Run in temporary git worktree |
**Cross-checks:**
- If `memory` set, agent should have Write access (auto-enabled by memory)
- Review-only agents: `disallowedTools: Write, Edit, NotebookEdit`
- `tools` and `disallowedTools` must not overlap
- Skills in `skills:` must exist in `plugins/elixir-phoenix/skills/`
**Detect changes:**
- Compare field list in docs against fields above
- Flag new fields the plugin doesn't use yet
- Flag fields the plugin uses that docs don't document (potential removal)
### Structural
- Valid markdown with YAML frontmatter (between `---` delimiters)
- Specialist agents: ≤365 lines
- Orchestrator agents (has `Agent` in tools): ≤535 lines
## Skill Validation Rules
### Structure
Check each `skills/*/` directory against `skills.md` docs.
**Required:**
- `SKILL.md` exists with `name` in frontmatter
**Frontmatter fields:**
| Field | Required | Notes |
|-------|----------|-------|
| `name` | Yes | Pattern: `phx:{name}` or `{domain}:{name}` |
| `description` | No | Used for auto-loading |
| `argument-hint` | No | Shown in command help |
| `disable-model-invocation` | No | Boolean, default false |
| `user-invocable` | No | Boolean, default true. Set false to hide from `/` menu |
| `allowed-tools` | No | Pre-approves tools while the skill is active (never restricts). Plugin skills must not set it |
| `model` | No | Model to use when skill is active |
| `context` | No | Set to `fork` to run in forked subagent |
| `agent` | No | Subagent type when `context: fork` is set |
| `hooks` | No | Lifecycle hooks scoped to this skill |
**Forbidden:** `triggers:` — MUST NOT be present.
**Detect changes:** New frontmatter fields, changed structure conventions.
### Structural
- SKILL.md: ≤185 lines
- references/*.md: ≤350 lines each
## Hook Validation Rules
### Schema
Check `hooks/hooks.json` against `hooks.md` docs.
**Structure:**
```json
{ "hooks": { "EventName": [{ "matcher": "", "hooks": [{ "type": "command", "command": "..." }] }] } }
```
**Valid event names:**
`PreToolUse`, `PostToolUse`, `PostToolUseFailure`, `PermissionRequest`,
`UserPromptSubmit`, `Notification`, `Stop`, `SubagentStart`, `SubagentStop`,
`SessionStart`, `SessionEnd`, `TeammateIdle`, `TaskCompleted`, `PreCompact`,
`InstructionsLoaded`, `ConfigChange`, `WorktreeCreate`, `WorktreeRemove`
**Valid hook types:**
| Type | Required Fields | Optional Fields |
|------|----------------|-----------------|
| `command` | `command` | `timeout`, `environment` |
| `prompt` | `prompt` | `model`, `tools` |
| `agent` | `prompt` | `model`, `tools`, `maxTurns` |
**Cross-checks:**
- Event names not in valid set = **BLOCKER** (silently ignored by Claude Code)
- `command` paths with `${CLAUDE_PLUGIN_ROOT}` should resolve to existing scripts
- Check docs for new event names, hook types, or fields
## Plugin Config Validation Rules
### plugin.json
**Required:** `name` (kebab-case, no spaces)
**Valid optional fields:**
`version`, `description`, `author` (`{name, email?, url?}`), `homepage`,
`repository`, `license`, `keywords`, `commands`, `agents`, `skills`,
`hooks`, `mcpServers`, `outputStyles`, `lspServers`
**Cross-checks:** All path fields resolve to existing files/directories.
### marketplace.json
**Required:** `name`, `owner` (object with `name`), `plugins` (array)
**Each plugin entry:** `name` (required), `source` (required — path to plugin dir).
Optional: `description`, `version`, `author`, `category`.
**Cross-checks:** `source` paths exist, each has `.claude-plugin/plugin.json`,
names are unique.
## Priority Classification
Workers MUST classify every finding:
| Level | Meaning | Example |
|-------|---------|---------|
| **BLOCKER** | Breaks with current Claude Code | Invalid event name, removed field |
| **WARNING** | Deprecated/discouraged | Old field name, deprecated pattern |
| **INFO** | New capability available | New hook event, new frontmatter field |
| **PASS** | Validates correctly | Field values match docs |
## Output Template
```markdown
# {Type} Validation Report
**Files checked**: {count}
**Documentation version**: {date fetched}
## Breaking Changes (BLOCKER)
- **{file}:{line}** — {description}
- Current: `{what plugin has}`
- Expected: `{what docs say}`
## Deprecations (WARNING)
- **{file}** — {description}
- Replacement: `{recommended alternative}`
## New Features Available (INFO)
- **{feature}** — {description}
- Docs reference: {section}
## Validation Passed
- {count} files checked, {count} fields validated
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization. | Stars: 470k | Updated | Claude CodeCodex | Verified | Install |
| academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file). | Stars: 197k | Updated | Claude CodeCodex | Unverified | Install |
| doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |