{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_7b87euuherhv","handle":"ico-mcp","url":"https://wellknown.network/agents/ico-mcp","links":{"self":"https://wellknown.network/agents/ico-mcp/record.json","html":"https://wellknown.network/agents/ico-mcp","markdown":"https://wellknown.network/agents/ico-mcp/record.md","api":"https://wellknown.network/api/v1/agents/ico-mcp","status":"https://wellknown.network/api/v1/agents/ico-mcp/status","claim":"https://wellknown.network/agents/ico-mcp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/ico-mcp/claim.json","badge":"https://wellknown.network/agents/ico-mcp/badge.svg","openapi":"https://wellknown.network/openapi.json","history":"https://wellknown.network/api/v1/agents/ico-mcp/history","tools":"https://wellknown.network/api/v1/agents/ico-mcp/tools"},"ard":{"identifier":"urn:air::server:ico-mcp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"ico-mcp","summary":"An MCP server on top of ICO: an agent writes typed operators, composes them into pipelines, runs and profiles them, and turns what works into reviewed code.","description":"# ico-mcp\n\n[![CI](https://img.shields.io/github/actions/workflow/status/apriori3d/ico-mcp/ci.yml?label=CI&logo=github)](https://github.com/apriori3d/ico-mcp/actions/workflows/ci.yml)\n![Python](https://img.shields.io/badge/Python-3.11%E2%80%933.12-blue?logo=python)\n[![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)\n\nAn MCP server on top of [ICO](https://github.com/apriori3d/ico) with which\nan agent builds data pipelines from its own operators. The agent writes\noperators, small typed Python functions, and composes them into pipelines\nas JSON specs; the server checks, runs and profiles them. Operators that\nproved themselves in runs become the domain's stable code through a pull\nrequest a human reviews, and the next task starts from them instead of a\nblank script.\n\n- **Work accumulates.** Each task adds operators and pipelines to the\n  domain's library instead of one-off scripts.\n- **Errors before the run.** Types are checked at every link before\n  anything runs, so a broken pipeline does not burn GPU time.\n- **Fewer tokens.** The agent finds operators as compact cards by input and\n  output type instead of reading code, runs saved pipelines by name, and\n  gets data back as paths; long runs go on in the background.\n- **Visible runs.** The Live Board shows the plan and the progress of every\n  node, the profiler where time and memory go, `sample_run` what flows out\n  of a node; a run can be cancelled between items.\n- **Local or GPU.** The same pipeline runs here or on a RunPod GPU: the job\n  carries the operators' code, model weights are pinned assets.\n\n```\nsearch_operators → validate → run → save_pipeline → crystallize → PR\n       a step is missing → propose_operator, propose_types\n```\n\n## How it is built\n\n```mermaid\nflowchart TB\n  A[\"agent\"] -- \"MCP tools\" --> S[\"ico-mcp server, Live Board\"]\n  S --> kernels\n  subgraph kernels[\"kernels\"]\n    direction LR\n    K[\"main: catalog, validate, plans\"] ~~~ J[\"one p…","publisher":null,"homepage":"https://github.com/apriori3d/ico-mcp","repository":"https://github.com/apriori3d/ico-mcp","version":"0.1.2","license":"Apache 2.0","protocols":["mcp"],"tags":["agents","ico","mcp","pipelines"],"pricing":null,"endpoints":[{"url":"pypi:ico-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.ci-cd","name":"CI/CD & Deploy","confidence":1,"provenance":"derived"},{"slug":"productivity.crm","name":"CRM & Sales","confidence":1,"provenance":"derived"}],"categories":["dev","productivity"],"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":"ico-mcp","registry":"pypi","observedAt":"2026-11-05T21:26:44.372Z","publishedAt":"2026-10-25T14:22:18.479708Z","latestVersion":"0.1.2"},"toolSurface":null,"endpointFacts":[]},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"ico-mcp","url":"https://pypi.org/project/ico-mcp/","firstSeenAt":"2026-10-24T17:27:38.499Z","fetchedAt":"2026-11-05T21:25:49.359Z","normalizedAt":"2026-11-05T21:25:49.359Z"}]},"firstSeenAt":"2026-10-24T17:27:38.499Z","updatedAt":"2026-11-05T21:26:44.372Z"}