blueprint-curate-docs

安装量: 46
排名: #16027

安装

npx skills add https://github.com/laurigates/claude-plugins --skill blueprint-curate-docs
/blueprint:curate-docs
Curate library or project documentation into ai_docs entries optimized for AI agents - concise, actionable, gotcha-aware context that fits in PRPs.
Usage
:
/blueprint:curate-docs [library-name]
or
/blueprint:curate-docs project:[pattern-name]
When to Use This Skill
Use this skill when...
Use alternative when...
Creating ai_docs for PRP context
Reading raw documentation for ad-hoc tasks
Documenting library patterns for reuse
One-time library usage
Building knowledge base for project
General library research
Context
ai_docs directory: !
find docs/blueprint -maxdepth 1 -name 'ai_docs' -type d
Existing library docs: !
find docs/blueprint/ai_docs/libraries -name "*.md" -type f
Existing project patterns: !
find docs/blueprint/ai_docs/project -name "*.md" -type f
Library in dependencies: !
grep -m1 "^$1[\":@=]" package.json pyproject.toml requirements.txt
Parameters
Parse
$ARGUMENTS
:
library-name
Name of library to document (e.g.,
redis
,
pydantic
)
Location:
docs/blueprint/ai_docs/libraries/[library-name].md
OR
project:[pattern-name]
for project patterns
Location:
docs/blueprint/ai_docs/project/[pattern-name].md
Execution
Execute complete documentation curation workflow:
Step 1: Determine target and check existing docs
Parse argument to determine if library or project pattern
Check if ai_docs entry already exists
If exists → Ask: Update or create new version?
Check project dependencies for library version
Step 2: Research and gather documentation
For
libraries
:
Find official documentation URL
Search for specific sections relevant to project use cases
Find known issues and gotchas (WebSearch: "{library} common issues", "{library} gotchas")
Extract key sections with WebFetch
For
project patterns
:
Search codebase for pattern implementations:
grep -r "{pattern}" src/
Identify where and how it's used
Document conventions and variations
Extract real code examples from project
Step 3: Extract key information
Use cases
How/why this library/pattern is used in project
Common operations
Most frequent uses
Patterns we use
Project-specific implementations (with file references)
Configuration
How it's configured in this project
Gotchas
Version-specific behaviors, common mistakes, performance pitfalls, security considerations Sources for gotchas: GitHub issues, Stack Overflow, team experience, official docs warnings. Step 4: Create ai_docs entry Generate file at appropriate location (see REFERENCE.md ): docs/blueprint/ai_docs/libraries/[library-name].md OR docs/blueprint/ai_docs/project/[pattern-name].md Include all sections from template: Quick Reference, Patterns We Use, Configuration, Gotchas, Testing, Examples. Keep under 200 lines total. Step 5: Add code examples Include copy-paste-ready code snippets from: Project codebase (reference actual files and line numbers) Official documentation examples Stack Overflow solutions Personal implementation experience Step 6: Update task registry Update the task registry entry in docs/blueprint/manifest.json : jq --arg now " $( date -u +%Y-%m-%dT%H:%M:%SZ ) " \ --argjson processed " ${ITEMS_PROCESSED :- 0} " \ --argjson created " ${ITEMS_CREATED :- 0} " \ '.task_registry["curate-docs"].last_completed_at = $now | .task_registry["curate-docs"].last_result = "success" | .task_registry["curate-docs"].stats.runs_total = ((.task_registry["curate-docs"].stats.runs_total // 0) + 1) | .task_registry["curate-docs"].stats.items_processed = $processed | .task_registry["curate-docs"].stats.items_created = $created' \ docs/blueprint/manifest.json

tmp.json && mv tmp.json docs/blueprint/manifest.json Step 7: Validate and save Verify entry is < 200 lines Verify all code examples are accurate Verify gotchas include solutions Save file Report completion Agentic Optimizations Context Command Check ai_docs exists test -d docs/blueprint/ai_docs && echo "YES" || echo "NO" List library docs ls docs/blueprint/ai_docs/libraries/ 2>/dev/null Check library version grep "{library}" package.json pyproject.toml 2>/dev/null | head -1 Search for patterns Use grep on src/ for project patterns Fast research Use WebSearch for common issues instead of fetching docs For ai_docs template, section guidelines, and example entries, see REFERENCE.md .

返回排行榜