Context7 Usage Skill

Trigger Words: context7, current docs, latest API, version-specific, use context7, query context7

What: Guide for using Context7 MCP with AIKnowSys workflows

When to use this skill:

•AI needs current library documentation
•Validating learned skills against latest APIs
•Scaffolding with specific framework version
•Planning with unfamiliar technologies
•Checking if patterns are still recommended

Prerequisites

•Context7 MCP installed and configured
•MCP-compatible AI client (Claude Desktop, Cursor, etc.)
•See docs/context7-integration.md for setup

Workflow

Step 1: Identify the Use Case

Ask yourself:

•Am I working with an external library/framework?
•Do I need current API documentation?
•Is this a version-specific question?

If YES → Use Context7
If NO → Use existing project knowledge

Step 2: Determine Library ID

Format: /owner/repo or /owner/repo/version

Common examples:

code

/vercel/next.js           # Latest Next.js
/vercel/next.js/v15.0.0   # Specific version
/facebook/react           # Latest React
/supabase/supabase        # Supabase
/mongodb/docs             # MongoDB
/vitejs/vite              # Vite

Finding IDs:

•Ask AI: "What's the Context7 library ID for [framework]?"
•Check Context7 docs: https://context7.com/libraries
•Format: GitHub repository path

Step 3: Formulate Query

Good queries (specific):

code

"Query Context7 for Next.js 15 middleware patterns"
"Use Context7: How to implement auth in Supabase v2?"
"Check Context7 for current React hooks best practices"

Poor queries (too vague):

code

"How does Next.js work?"  # Too broad
"Get docs"  # Missing library
"Latest API"  # Which library?

Best practice: Include library name + version + specific topic

Step 4: Apply Response to AIKnowSys

For learned skills:

markdown

# Skill: Next.js Middleware Pattern

**Source:** Context7 `/vercel/next.js/v15.0.0` (Feb 2026)

**Pattern:**
\`\`\`typescript
// Current Next.js 15 middleware
export function middleware(request: NextRequest) {
  // ... (code from Context7)
}
\`\`\`

**When to use:** [your project-specific context]

For tech stack scaffolding:

•Use Context7 response to verify CODEBASE_ESSENTIALS patterns
•Update tech stack examples with current conventions
•Reference version in documentation

For validation:

•Compare existing skill with Context7 response
•Flag deprecated patterns
•Suggest updates if drift detected

Integration Points

With @Planner Agent

Scenario: Planning a new feature with unfamiliar tech

Workflow:

code

1. @Planner I need to implement [feature] with [technology]
2. AI: "I'll query Context7 for current patterns"
3. AI uses Context7 to get latest docs
4. Planner creates plan with accurate APIs
5. Plan includes version references

Example:

code

@Planner Plan authentication for Next.js 15 app
Use Context7 to get current middleware and route handler patterns.

With @Developer Agent

Scenario: Implementing a feature with external dependency

Workflow:

code

1. Developer reads CODEBASE_ESSENTIALS (project patterns)
2. Identifies external library usage
3. Queries Context7 for current API
4. Implements with up-to-date patterns
5. Documents in learned skill with version

Example:

code

@Developer Implement Supabase auth
Use Context7 for Supabase v2 current auth patterns.
Follow TDD workflow.

With Skill Creation

Scenario: Creating a learned skill about external library

Workflow:

code

1. Discover pattern in your code
2. Query Context7 for library documentation
3. Validate pattern is current best practice
4. Document in SKILL.md with:
   - Source: Context7 /owner/repo/version
   - Last validated: Date
   - Version-specific notes

Example skill header:

markdown

# Skill: Supabase Row Level Security

**Source:** Context7 `/supabase/supabase` (Feb 2026)
**Last Validated:** 2026-02-01
**Supabase Version:** 2.x

Best Practices

DO ✅

•Include version numbers in queries and skill documentation
•Validate before committing learned skills
•Cache responses for repeated queries (24 hour TTL)
•Document source when using Context7 in skills
•Combine with project knowledge (Context7 + CODEBASE_ESSENTIALS)

DON'T ❌

•Don't use for project-specific code (use CODEBASE_ESSENTIALS)
•Don't skip version when version matters
•Don't trust blindly - validate Context7 responses make sense
•Don't query for general concepts (use built-in AI knowledge)
•Don't use for performance-critical workflows (adds latency)

When Context7 is Unavailable

Graceful degradation:

code

If Context7 not available:
  1. Use AI's built-in knowledge (may be outdated)
  2. Manually check official docs
  3. Add TODO to validate with Context7 later
  4. Document assumption in code

Example:

javascript

// TODO: Validate with Context7 when available
// Assumption: Next.js 15 middleware pattern (as of Feb 2026)
export function middleware(req) {
  // ...
}

Troubleshooting

Query Returns No Results

Possible causes:

•Library ID incorrect (check format)
•Version doesn't exist or isn't documented
•Library not indexed by Context7

Solutions:

•Verify library ID with AI: "Is /owner/repo correct?"
•Try without version (use latest)
•Check Context7 library list: https://context7.com/libraries
•Manual fallback: Check official docs directly

Response Seems Outdated

Possible causes:

•Context7 cache
•Library docs not updated
•Wrong version specified

Solutions:

•Clear cache: .aiknowsys/context7-cache/
•Verify version number matches expectation
•Check library's official changelog
•Report to Context7 if docs should be current

Timeout or Slow

Possible causes:

•Network latency
•Large documentation
•Rate limiting

Solutions:

•Be more specific in query (reduce scope)
•Enable caching (see Advanced Features in docs)
•Use cached skills for repeated needs
•Consider offline fallback workflow

Examples

Example 1: Validating Learned Skill

Scenario: You created a skill 3 months ago about React hooks

Workflow:

code

1. Open skill: .github/skills/react-hooks-pattern/SKILL.md
2. Ask AI: "Use Context7 to check if this React hooks skill is current.
            Compare with /facebook/react latest docs."
3. AI queries Context7
4. AI reports: "✅ Pattern is current" or "⚠️ Deprecated, use X instead"
5. Update skill if needed
6. Update "Last Validated" date

Example 2: Scaffolding with Specific Version

Scenario: Starting a new Next.js 15 project

Workflow:

bash

npx aiknowsys init --stack nextjs

# In AI chat:
"Use Context7 /vercel/next.js/v15.0.0 to verify:
 1. App Router structure is current
 2. Middleware patterns are up-to-date
 3. Server Components conventions are correct
 
Update CODEBASE_ESSENTIALS if any patterns are outdated."

Example 3: Planning with Unfamiliar Tech

Scenario: Need to add Supabase to existing project

Workflow:

code

@Planner Create plan for Supabase integration
Requirements:
- Authentication (email/password + OAuth)
- Row Level Security
- Real-time subscriptions

Use Context7 /supabase/supabase to:
1. Get current auth patterns
2. Check RLS best practices
3. Verify real-time API

Create implementation plan with current APIs.

Related Skills

•skill-validation - Automated skill validation
•skill-creator - Creating learned skills
•tdd-workflow - Test-driven development

Context7 Usage Skill

Prerequisites

Workflow

Step 1: Identify the Use Case

Step 2: Determine Library ID

Step 3: Formulate Query

Step 4: Apply Response to AIKnowSys

Integration Points

With @Planner Agent

With @Developer Agent

With Skill Creation

Best Practices

DO ✅

DON'T ❌

When Context7 is Unavailable

Troubleshooting

Query Returns No Results

Response Seems Outdated

Timeout or Slow

Examples

Example 1: Validating Learned Skill

Example 2: Scaffolding with Specific Version

Example 3: Planning with Unfamiliar Tech

Related Skills

See Also