generated: '2026-09-23' method: probed source: https://buildwithfern.com/learn/_mcp/server status: published server: name: fern-api transport: http url: https://buildwithfern.com/learn/_mcp/server deployment: mode: remote endpoint: https://buildwithfern.com/learn/_mcp/server verified: probed probe: live checked: '2026-09-23' source: docs-check 2026-09-23 (MCP server named in the provider's own llms.txt/docs, then probed; roadmap#590) tools_note: No tool list captured. This manifest records a PROBED, reachable endpoint only; anonymous tools/list was auth-gated or the server requires an initialize handshake this reachability probe does not perform. Run signals/audit_mcp_authorship.py --probe to enumerate tools and settle platform-vs-provider authorship. Do NOT derive a tool list from the OpenAPI and store it here -- that is a candidate manifest, and it belongs under an X-MCPServerCandidate pointer. tool_quality: tool_count: 2 tokens_total: 302 tokens_per_tool_median: 151 heavy_tools: [] heavy_tool_count: 0 params_total: 3 params_undescribed: 0 params_described_pct: 100.0 tools_with_undescribed_params: 0 tools_annotated: 2 annotated_pct: 100.0 state_changing_unhinted: [] tools_missing_schema: 0 short_descriptions: 0 duplicate_descriptions: 0 duplicate_names: 0 nonportable_names: [] output_schema_count: 0 token_estimate: chars/3.6 of each compact tool definition source: mcp/fern-api-mcp-tools.json measured: '2026-10-20' scored: false note: Facts from the saved tools/list capture (mcp_tool_quality.py); recorded, not scored. Checks learned from Glimind's agent-readiness audit. tools: - name: searchDocs description: Search the documentation at https://buildwithfern.com/learn. Returns ranked results with title, url, section anchor and a snippet, in rank order. Use fetchPage to read a result in full. - name: fetchPage description: Read one page of the documentation at https://buildwithfern.com/learn as markdown. Takes a path or url from a searchDocs result. tools_count: 2