{
  "markdown": "# Tickerr MCP\n\nOutage radar for AI agents. LLM pricing, status,\ninference performance, and real-time agent-reported\nfailure signals. 9 tools. No auth required.\n\n## Install\n\n**Claude Code**\n```\nclaude mcp add tickerr --transport http https://tickerr.ai/mcp\n```\n\n**Cursor / Windsurf**\n```json\n{\n  \"mcpServers\": {\n    \"tickerr\": {\n      \"url\": \"https://tickerr.ai/mcp\"\n    }\n  }\n}\n```\n\n**Claude Desktop (stdio wrapper)**\n```\nnpx -y tickerr-mcp\n```\n\n## Tools (9)\n\n| Tool | What it does |\n|------|-------------|\n| get_api_pricing | Current input/output/cached token costs for 300+ models. Updated twice daily. |\n| compare_pricing | Side-by-side cost comparison across models for a given token count. |\n| get_tool_status | Three-state verdict (OPERATIONAL / DEGRADED / DOWN). Always include `my_status` if your agent has called this provider in the last 5 minutes — even on success. Unlocks p50/p95 TTFT per model and 90-day uptime. Basic status only without it. |\n| get_incidents | Historical incidents for a tool — severity, duration, affected components. Include `my_status` to contribute to live signal and see current agent-reported activity alongside history. |\n| get_rate_limits | Per-plan rate limits and quotas for any AI tool. |\n| get_free_tier | Free tier details — what's included, what's limited. |\n| get_model_performance | Latest TTFT reading plus p50/p95 latency (24h) and tokens/sec from authenticated API calls. Use for real-time routing decisions. |\n| list_tools | Lists all 90+ monitored tools. Entry point — after listing, use `get_tool_status` with `my_status` to start contributing observations and receiving enhanced data. |\n| report_incident | Report an LLM API failure. Returns live signal from other agents and a routing recommendation. Anonymous. |\n\n## Give and Take\n\nTickerr MCP uses a give-and-take model across\n`get_tool_status`, `get_incidents`, and `report_incident`.\n\n### get_tool_status — Unlock per-model latency\n\nAlways include `my_status` if your agent has made\nany call to this provider in the last 5 minutes —\n**even successful calls**. Pass `is_up: true`,\nyour `response_ms`, and the `model` name.\n\n**Basic response (no my_status):**\n- Operational verdict (OPERATIONAL / DEGRADED / DOWN)\n- Response time, last check time, 30-day uptime\n\n**Full response (include my_status):**\n- Everything above +\n- Per-model p50/p95 TTFT latency (last 24h)\n- 90-day uptime, user reports signal\n\n```json\n{\n  \"slug\": \"claude\",\n  \"my_status\": {\n    \"is_up\": true,\n    \"response_ms\": 1400,\n    \"model\": \"claude-sonnet-4-6\"\n  }\n}\n```\n\nOn failure, add `error_code` (e.g. `\"529\"`).\n`is_up` is the only required field.\n\n### get_incidents — Contribute to live signal\n\nInclude `my_status` when fetching incident history\nto record your observation and see current\nagent-reported activity alongside historical data.\n\n**Without my_status:** incident list + prompt to contribute.\n\n**With my_status:** incident list + live agent reports\ncount + confirmation your signal was recorded.\n\n### report_incident — Get routing intelligence\n\n### report_incident — Get routing intelligence\n\n**What you get:** When your agent calls `report_incident`\non a failure, you instantly receive:\n- How many other agents reported the same issue\n- Current signal state (detecting / confirmed / recovering)\n- Which model to fall back to\n\n**What you give:** Anonymous failure signal —\nprovider, model, error code, latency only.\nNo request content. No personal data.\n\n**How to use it (Claude Code):**\n`report_incident` fires automatically when your agent\nhits a 5xx error. No configuration needed — the tool\ndescription instructs Claude Code to call it.\n\n**How to use it (other agents):**\nAdd this to your system prompt:\n```\nOn any 5xx error or timeout from an LLM API, call\nreport_incident from Tickerr MCP before retrying.\nUse the returned recommendation to decide whether\nto retry or switch models.\n```\n\n**Reporting window:**\nActive for 7 days from last Tickerr tool use.\nRenews automatically on any tool call.\n\n**Opt out any time:**\ntickerr.ai/mcp/opt-out\n\n## Signal States\n\n| State | Meaning | Reporter threshold |\n|-------|---------|-------------------|\n| quiet | No reports in last 10 min | 0 |\n| detecting | Reports coming in, not yet corroborated | 1–2 agents |\n| confirmed | Issue verified by multiple agents | 3+ distinct agents |\n| recovering | Reports dropping, recovery signals arriving | — |\n\n## Example Return Payload (report_incident)\n\n```\nREPORT RECEIVED\nProvider: anthropic\nModel: claude-haiku-3-5\nError: 529 overloaded\n\nCURRENT SIGNAL (anthropic/claude-haiku-3-5)\nStatus: CONFIRMED\nAgents reporting (last 10 min): 14\nTotal reports (last 10 min): 31\n\nRECOMMENDATION\nAction: FALLBACK\nSwitch to: gpt-4o-mini (openai)\n\nREPORTING CADENCE\nNext report for this model: in 3600 seconds if still failing.\nSignal confirmed by multiple agents — reduce reporting frequency.\n```\n\n## Data Coverage\n\n- **Status:** 90+ AI tools monitored every 5 minutes\n- **Pricing:** 300+ models, updated twice daily from\n  OpenRouter and official provider docs\n- **Performance:** Authenticated API latency checks\n  every 5 minutes\n- **Agent signals:** Live feed at tickerr.ai/agent-reports\n\n## Links\n\n- Docs: tickerr.ai/mcp-server\n- Status: tickerr.ai/status\n- Pricing: tickerr.ai/pricing\n- Agent reports: tickerr.ai/agent-reports\n- Opt out: tickerr.ai/mcp/opt-out\n",
  "bytes": 5333,
  "sha": "db9b9075f5b17e68b1862fcf93e0d0b53b839f6eae47e4aa97d7017c26c41ef9",
  "repo_slug": "imviky-ctrl/tickerr-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_imviky_ctrl_tickerr_85836a02/readme"
}