knowledge-router_skill

This skill decides when to route questions between curated skills and Context7 to optimize documentation lookups for frameworks, libraries, and APIs.
  • Python

4

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill jawwad-ali/claude-code-skills --skill knowledge-router

  • SKILL.md8.8 KB

Overview

This skill routes technical documentation and pattern requests between curated skills and the Context7 MCP documentation service. It decides whether to answer from a validated pattern, fetch authoritative docs, or combine both for complex tasks to produce accurate, practical guidance. Use it to optimize retrieval strategy for frameworks, libraries, and APIs.

How this skill works

The router analyzes the user's query to classify it as a pattern request, a lookup request, or a hybrid need. For common implementation tasks it selects a curated skill; for precise API/syntax/version questions it queries Context7; for complex integrations or debugging it loads a skill for architecture and then fetches specific docs from Context7 to fill gaps. It then synthesizes the best response based on that route.

When to use it

  • Answering common implementation or best-practice questions (CRUD, auth, routing).
  • Resolving exact API signatures, version changes, deprecated features, or pasted error messages.
  • Handling complex integrations that need both architecture patterns and precise API details.
  • Debugging sessions where pattern context plus specific error lookups are required.
  • Multi-turn conversations about the same technology where efficiency matters.

Best practices

  • Classify queries by intent keywords (e.g., "how do I", "what parameters", "latest").
  • Prefer curated skill for patterns and opinions; prefer Context7 for up-to-date, authoritative references.
  • For complex tasks, use the hybrid flow: generate pattern first, then query Context7 for missing specifics.
  • Load the relevant curated skill once for multi-turn flows to save tokens and maintain context.
  • When uncertain, escalate to hybrid rather than relying solely on one source.

Example use cases

  • "How do I add JWT authentication to FastAPI?" → use curated FastAPI skill for pattern and code samples.
  • "What parameters does HTTPException accept in FastAPI 0.100+?" → query Context7 for exact API reference.
  • "Build a production WebSocket chat with FastAPI and Redis" → hybrid: skill for architecture, Context7 for connection and scaling details.
  • Debugging: paste a stack trace → Context7 to find matching docs or issue tracker entries, then synthesize fix with skill patterns.
  • Integrations: combine pattern from one skill with Context7 docs for third-party library edge cases.

FAQ

Decisions use query signals: pattern keywords steer to skills; version, error, and API-precision steer to Context7; mixed signals trigger hybrid routing.

What if a curated skill is outdated?

The router will escalate to hybrid or Context7 when it detects version-specific or deprecated features, ensuring authoritative, current details.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational