SigRank MCP Server
15 tools any AI agent can call — measure, rank, and improve token efficiency. The yield cascade and live leaderboard as MCP tools.
Install
npx sigrankOr npx sigrank-mcp — same thing. No API key required for read tools. Submit tools need a Supabase anon key (free, set during sigrank enroll).
Client setup
| Client | Where |
|---|---|
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Cursor | Settings → MCP → Add Server |
| Cline | Marketplace → search "SigRank" |
| Windsurf | Settings → MCP Servers → Add |
| Any MCP client | npx sigrank (stdio transport) |
Tools (15)
rank_pastePaste token counts, get projected SigRank rank in secondsget_leaderboardPull the live global leaderboard from signalaf.comget_operatorGet any operator's profile, class tier, and cascade breakdownsubmit_pasteSubmit token telemetry to the live board (paste mode)submit_verifiedSubmit ed25519-signed telemetry (verified mode)tokenpullPull telemetry from ccusage, tokscale, or tokendashtokenpull_submitPull telemetry and submit in one callwatch_tokenpullContinuous monitoring with periodic submissionsrank_windowsCompare rank across 7d / 30d / 90d / all-time windowstokenpull_compareHead-to-head operator comparisonenrollRegister as an operator on the leaderboardsimulate_changeModel what-if scenarios (what if cache hit rate +10%?)diagnose_cascadeIdentify inefficiencies in token usage patternssuggest_improvementsGet actionable recommendations to improve yieldself_improveScope improvements to the agent's own telemetry