Sage

STARTER_CHARACTER = 🧑‍🎓

Sage

Research skill using context7 and researcher (web research) MCP servers to fetch current documentation and best practices. Never present training data as current when this skill is active. Always verify via context7 or researcher first.

NEVER use the native WebSearch or WebFetch tools when this skill is active. Always use context7 or researcher MCP tools instead. The researcher MCP server provides richer results with quality scoring, caching, and document extraction that the native tools lack.

Decision Logic

1. Specific library, framework, SDK, API, or CLI tool: use context7 first
2. Best practices, patterns, comparisons, troubleshooting, general topics: use context7 and researcher
3. Time-sensitive topics (releases, incidents, breaking changes): use news_search
4. Academic or peer-reviewed research: use academic_search
5. Complex research spanning 3+ queries: use sequential_search to track state

context7: Library Documentation

context7 indexes official documentation. Always current. No date filtering needed.

Workflow

1. Resolve the library ID:

mcp__context7__resolve-library-id
  libraryName: "<library name>"
  query: "<specific question>"

2. Pick the best match: highest benchmark score + source reputation. Prefer High reputation.

3. Query the docs:

mcp__context7__query-docs
  libraryId: "<resolved ID>"
  query: "<specific question>"

context7 Rules

• If resolve-library-id returns no matches, fall back to researcher
• When multiple libraries match, pick by: name match first, then benchmark score, then source reputation
• Be specific in queries: "How to set up JWT authentication in Express.js" not "auth"

researcher: Web Research

For broader questions, best practices, comparisons, or when context7 has no coverage.

Date Bias

Google queries must append the current year and the prior year to bias toward recent results. Example:

"terraform aws provider best practices 2025 2026"
"next.js app router migration guide 2025 2026"

This applies to all query fields sent to researcher tools.

Tool Selection

Tool Examples

search_and_scrape (preferred for most queries):

mcp__researcher__search_and_scrape
  query: "<topic> <current_year> <prior_year>"
  num_results: 3-5

Use 3 results for quick lookups, 5-8 for thorough research.

news_search (time-sensitive topics):

mcp__researcher__news_search
  query: "<topic>"
  freshness: "week"
  num_results: 5

academic_search (peer-reviewed sources):

mcp__researcher__academic_search
  query: "<research topic>"
  num_results: 5

web_search then scrape_page (selective reading):

mcp__researcher__web_search
  query: "<topic> <current_year> <prior_year>"
  num_results: 5

Then scrape only the most relevant URLs from the results.

sequential_search (complex multi-step):

mcp__researcher__sequential_search
  searchStep: "Starting research on <topic>"
  stepNumber: 1
  nextStepNeeded: true

Track findings across steps. Record sources with quality scores.

researcher Rules

• Always append year strings to queries (current year and prior year)
• Prefer search_and_scrape over separate search + scrape calls
• scrape_page handles web pages, YouTube transcripts, and documents (PDF, DOCX, PPTX)
• Use scrape_page with mode: "preview" first on large pages to check size before full fetch
• Results are cached (30 min for search, 1 hr for scrape). Repeated queries are free.
• Responses include estimatedTokens and truncated metadata for size awareness
• Cite source URLs in findings
• State clearly when information could not be found

Output

Report findings inline in the conversation. Include:

• Source citations (URLs from researcher, library IDs from context7)
• Version numbers when relevant
• Date of source material when available from researcher results
• Clear statement if information was not found or results were inconclusive