{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_zngncdsu67nc","handle":"openinterp-mcp","url":"https://wellknown.network/agents/openinterp-mcp","links":{"self":"https://wellknown.network/agents/openinterp-mcp/record.json","html":"https://wellknown.network/agents/openinterp-mcp","markdown":"https://wellknown.network/agents/openinterp-mcp/record.md","api":"https://wellknown.network/api/v1/agents/openinterp-mcp","status":"https://wellknown.network/api/v1/agents/openinterp-mcp/status","claim":"https://wellknown.network/agents/openinterp-mcp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/openinterp-mcp/claim.json","badge":"https://wellknown.network/agents/openinterp-mcp/badge.svg","openapi":"https://wellknown.network/openapi.json","history":"https://wellknown.network/api/v1/agents/openinterp-mcp/history","tools":"https://wellknown.network/api/v1/agents/openinterp-mcp/tools"},"ard":{"identifier":"urn:air::server:openinterp-mcp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"openinterp-mcp","summary":"MCP server + Colab backend for mechanistic interpretability research. Bring-your-own-agent. Privacy-first.","description":"# openinterp-mcp\n\n> **v0.1.0 beta · API may shift before v1.0**\n>\n> MCP server + Colab backend for mechanistic interpretability research.\n> **Bring your own agent.** Works with Claude Code, Cursor, Cline, OpenHands, Aider, or any harness that speaks MCP.\n> **Privacy-first.** We do not host inference. We do not custody your keys. Your Colab session, your model, your data.\n\n<a href=\"https://www.producthunt.com/products/openinterpretability?embed=true&utm_source=badge-featured&utm_medium=badge&utm_campaign=badge-openinterpretability\" target=\"_blank\"><img src=\"https://api.producthunt.com/widgets/embed-image/v1/featured.svg?post_id=1146555&theme=light&t=1778728126905\" alt=\"OpenInterpretability - Open-source toolkit to audit what your LLM knows | Product Hunt\" width=\"250\" height=\"54\" /></a>\n\n## What this is\n\nA research toolkit that turns probe-causality and SAE-feature experiments into agent-callable primitives. Researchers run the model on their own compute (Colab Pro recommended), expose an HTTPS endpoint over ngrok, and let an LLM agent drive experiments via **8 typed MCP tools**.\n\nThe 8 MCP primitives:\n\n| Tool | What it does |\n|---|---|\n| `colab_attach` | Attach to a running Colab/vast.ai/runpod session via its public HTTPS URL. Validates `/health`, caches the endpoint locally. |\n| `colab_status` | Health check — loaded model, probes in memory, captures held. |\n| `list_probes` | List probes currently loaded in the backend (model_id, layer, position, source). |\n| `capture_acts` | Run a forward pass with hooks, extract activations at specified layers/positions. Returns `capture_id`. |\n| `probe_eval` | Apply a loaded probe to a stored capture, return AUROC + per-sample scores. |\n| `steer` | Inject direction×α at layer L. Returns base + steered generation + control-token-normalized Δrel (paper-6 protocol). |\n| `sae_lookup` | Decompose a stored activation into top-K SAE features with auto-interp descriptions. |\n| `causality_protocol` | Run the three mandatory checks (rand…","publisher":null,"homepage":"https://openinterp.org/mcp","repository":"https://github.com/OpenInterpretability/openinterp-mcp/issues","version":"0.2.0","license":"Apache-2.0","protocols":["mcp"],"tags":["mechanistic-interpretability","mcp","claude-code","probes","sae","agent-monitoring"],"pricing":null,"endpoints":[{"url":"pypi:openinterp-mcp","type":"package_pypi","auth":null,"probeable":false}],"skills":null,"tools":null,"extra":null,"attribution":{"kind":"pypi","name":"pypi","license":"pypi","repoUrl":"pypi","summary":"pypi","version":"pypi","description":"pypi","homepageUrl":"pypi"}},"derived":{"capabilities":[{"slug":"dev.monitoring","name":"Monitoring & Observability","confidence":0.713,"provenance":"derived"}],"categories":["dev"],"language":"en"},"observed":{"status":"unknown","statusReason":"Distributed as a package to run locally; no network endpoint to check.","lastOkAt":null,"lastProbedAt":null,"statusComputedAt":null,"reliability30d":null,"latestObservations":[],"tools":null,"package":{"name":"openinterp-mcp","registry":"pypi","observedAt":"2026-11-03T20:29:47.585Z","publishedAt":"2026-06-17T12:17:47.202797Z","latestVersion":"0.2.0"},"toolSurface":null,"endpointFacts":[]},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"openinterp-mcp","url":"https://pypi.org/project/openinterp-mcp/","firstSeenAt":"2026-09-10T08:26:46.467Z","fetchedAt":"2026-11-03T20:29:05.145Z","normalizedAt":"2026-11-03T20:29:05.145Z"}]},"firstSeenAt":"2026-09-10T08:26:46.467Z","updatedAt":"2026-11-03T20:29:47.585Z"}