Skip to content
Back to search

overmind

MarketplaceWorks with Claude CodeGitHub

Overmind plugins for coding agents.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/overmind-core/overmind with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/overmind-core/overmind'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/overmind-core--overmind/badge.svg)](/listings/overmind-core--overmind)

HTML

<a href="/listings/overmind-core--overmind"><img src="/api/listings/overmind-core--overmind/badge.svg" alt="findsafeskills safety rating" /></a>

Install

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 15 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "name": "overmind",
  3. "owner": {
  4. "name": "Overmind",
  5. "email": "hello@overmindlab.ai"
  6. },
  7. "description": "Overmind plugins for coding agents.",
  8. "plugins": [
  9. {
  10. "name": "overmind",
  11. "source": "./overmind",
  12. "description": "Overmind skills — agent observability, tracing, and platform workflows for coding agents."
  13. }
  14. ]
  15. }
README.md · README · 243 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. <img width="6000" height="2000" alt="X Company Banner Stone (1)" src="https://github.com/user-attachments/assets/8ba6a64f-0819-47bd-9d58-af89ee3e7bad" />
  2. # The Training Platform for Specialized Agents
  3. <p align="center">
  4. <a href="https://console.overmindlab.ai/?utm_source=github&amp;utm_medium=readme&amp;utm_campaign=always-on&amp;utm_content=readme-header">Console</a> | <a href="https://www.overmindlab.ai/?utm_source=github&amp;utm_medium=readme&amp;utm_campaign=always-on&amp;utm_content=readme-header">Site</a> | <a href="#run-it-yourself">Self-host</a>
  5. </p>
  6. <p align="center">
  7. <a href="https://docs.overmindlab.ai"><img src="https://img.shields.io/badge/Docs-ed670f?style=for-the-badge" alt="Documentation"></a>
  8. <a href="https://discord.gg/TPF722ZKuj"><img src="https://img.shields.io/badge/Discord-5865F2?style=for-the-badge&logo=discord&logoColor=white" alt="Discord"></a>
  9. <a href="https://pypi.org/project/overmind/"><img src="https://img.shields.io/pypi/v/overmind?style=for-the-badge&label=PyPI&color=3b1b06" alt="PyPI"></a>
  10. <a href="https://github.com/overmind-core/overmind/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/overmind-core/overmind/ci.yml?style=for-the-badge&label=CI" alt="CI"></a>
  11. <a href="LICENSE"><img src="https://img.shields.io/badge/Platform-AGPL--3.0-3b1b06?style=for-the-badge" alt="Platform AGPL-3.0"></a>
  12. <a href="overmind/LICENSE"><img src="https://img.shields.io/badge/SDK-MIT-3b1b06?style=for-the-badge" alt="SDK MIT"></a>
  13. </p>
  14. **Overmind continuously trains & improves your agents, with data from your production traces**
  15. The SDK and CLI you install (`pip install overmind`) are MIT. The platform behind them is AGPL-3.0 and you can self-host it. Details in [Licence](#licence).
  16. Point it at your agent's codebase and it turns production traces (or any dataset) into a fine-tuned model, benchmarked against the eval metrics you define and served via 1 unified API, with no ML infrastructure to build.
  17. > The weights are yours to download, retrain or roll back.
  18. Available from the [Console](https://console.overmindlab.ai/?utm_source=github&utm_medium=readme&utm_campaign=always-on&utm_content=readme-intro), the `overmind` CLI, the [REST API](https://docs.overmindlab.ai/reference/api), and an [MCP server](#connect-your-coding-agent) for Cursor, Claude Code, OpenCode and Codex. Hosted at [console.overmindlab.ai](https://console.overmindlab.ai/?utm_source=github&utm_medium=readme&utm_campaign=always-on&utm_content=readme-intro) or [run it yourself](#run-it-yourself).
  19. <table>
  20. <tr><td><b><a href="https://docs.overmindlab.ai/explanation/capabilities">Agent & Capabilities</a></b></td><td>A graph of your agent — capabilities, prompts, tools, and tasks — scanned from the repo.</td></tr>
  21. <tr><td><b><a href="https://docs.overmindlab.ai/how-to/investigate">Observability</a></b></td><td>OpenTelemetry traces, scored as they arrive and matched to the capability that produced them.</td></tr>
  22. <tr><td><b><a href="https://docs.overmindlab.ai/how-to/prepare-data">Datasets</a></b></td><td>Production traces or uploaded files become versioned eval and training datasets.</td></tr>
  23. <tr><td><b><a href="https://docs.overmindlab.ai/how-to/evaluate">Eval</a></b></td><td>What "good" means per capability, measured on live traces and in batch.</td></tr>
  24. <tr><td><b><a href="https://docs.overmindlab.ai/how-to/optimise">Optimisers</a></b></td><td>Prompt, tool, and control-flow experiments in your repo; the winner is a git diff.</td></tr>
  25. <tr><td><b><a href="https://docs.overmindlab.ai/how-to/train">Models</a></b></td><td>Fine-tunes you own, trained on your data and benchmarked against production.</td></tr>
  26. <tr><td><b><a href="https://docs.overmindlab.ai/how-to/serve">Inference</a></b></td><td>Trained and frontier models on one OpenAI-compatible API.</td></tr>
  27. </table>
  28. <p align="center">
  29. <a href="https://youtu.be/DWC0BO48154">
  30. <img width="9872" height="5543" alt="playframe" src="https://github.com/user-attachments/assets/246a21c5-e07a-4414-981e-f60d6308a729" />
  31. </a>
  32. </p>
  33. ## Get started
  34. ### Hosted
  35. Sign up at [console.overmindlab.ai](https://console.overmindlab.ai/?utm_source=github&utm_medium=readme&utm_campaign=always-on&utm_content=readme-get-started), copy the onboarding prompt from **Get started**, and paste it into your coding agent (Cursor, Claude Code, OpenCode or Codex) with your agent's repo root open. The **Data** tab starts from an uploaded dataset instead. It installs `overmind`, runs `overmind init` and `overmind sync`, and maps the agent's capabilities. From then on everything is a `/overmind` command in the same chat:
  36. | Command | What it does |
  37. | -------------------------- | --------------------------------------------- |
  38. | `/overmind ensure-tracing` | Inspect traces and instrument the agent |
  39. | `/overmind dataset` | Build, clean, upload or export a dataset |
  40. | `/overmind finetune` | Fine-tune, deploy and smoke-test a model |
  41. | `/overmind optimise` | Run prompt and code optimisation |
  42. | `/overmind backtest` | Compare models against the agent's own traces |
  43. ### Run it yourself
  44. Self-hosting keeps traces and training data inside your own network. The hosted and self-hosted stacks are the same code.
  45. ```bash
  46. git clone https://github.com/overmind-core/overmind.git && cd overmind
  47. cp .env.example .env # set the required keys below
  48. docker compose up -d # Postgres, Redis, API on :8000, Console on :5173, Celery workers, beat, Workshop sandbox, Grafana on :3001
  49. ```
  50. Document uploads include local English OCR for scanned PDFs, embedded images and direct PNG/JPEG/WebP uploads. Direct images are capped at 100 MB and 64 megapixels; animated images are not supported. The Docker image includes Tesseract and its English/orientation data. For workers or backend tests outside Docker, install `tesseract-ocr`, `tesseract-ocr-eng` and `tesseract-ocr-osd` on Debian/Ubuntu, or `brew install tesseract` on macOS. No external OCR service is required.
  51. On first boot the API runs migrations; Swagger is at `/api/docs/`. Sign in at `http://localhost:5173` with any email and password. `docker compose exec api python manage.py seed_demo --owner <your email>` loads a full demo workspace.
  52. <details>
  53. <summary><b>What the API needs to boot</b></summary>
  54. The API refuses to start until every required key is set, and the error names each missing one. `.env.example` documents every key.
  55. | Required | Used for |
  56. | --------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------- |
  57. | `OPENROUTER_API_KEY` | Judges, evals, trace scoring, optimiser scoring, frontier inference |
  58. | `MODAL_TOKEN_ID`, `MODAL_TOKEN_SECRET` | Modal training and serving workers (set `MODAL_ENVIRONMENT` explicitly) |
  59. | `INFERENCE_API_URL`, `INFERENCE_API_KEY` | Serving trained models — the endpoint printed by `modal deploy` and its shared secret |
  60. | `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`, `AWS_BUCKET_NAME` | The fine-tuning checkpoint archive |
  61. | `HF_TOKEN` | Gated Hugging Face base models; also set it in the Modal secret |
  62. | `SANDBOX_URL`, `SANDBOX_TOKEN` | The isolated Workshop sandbox; `.env.example` and compose set local values — change the token in production |
  63. | Optional | Effect |
  64. | ------------------- | ----------------------------------------------------------------------- |
  65. | `STRIPE_SECRET_KEY` | Paid plans and the credit cap; without it, usage is metered with no cap |
  66. Fine-tuning and serving also need the Modal workers deployed (`modal deploy overbae/modal/modal_vllm_worker.py`, `register_model.py` and `modal_sft_worker.py`; native decision models also use `modal_decision_evaluation.py` and `modal_decision_serving.py`) and a Modal secret named `overmind-inference` with the AWS keys, `INFERENCE_API_KEY` and `HF_TOKEN`.
  67. Self-hosted login uses email and password. Keep Clerk and Stripe secrets blank.
  68. Data Workshop transformations are authored by your native coding agent through MCP;
  69. the platform stores source files, immutable cell versions and execution receipts.
  70. There is no platform Workshop agent or ChatGPT account connection.
  71. Reusable scripts, cross-source revisions and continuous bindings use the dedicated
  72. [Workshop runtime and package protocol](https://docs.overmindlab.ai/reference/pipelines).
  73. See [self-hosting](https://docs.overmindlab.ai/how-to/self-host) for configuration
  74. and private VM setup.
  75. </details>
  76. ### Send a first trace
  77. ```bash
  78. pip install "overmind[tracing]"
  79. export OVERMIND_API_KEY=ovr_… # project key from the project page in the Console; add OVERMIND_API_URL for self-host
  80. ```
  81. ```python
  82. import overmind
  83. overmind.init(
  84. service_name="support-agent", capability_id="<capability-uuid>", providers="auto"
  85. )
  86. @overmind.tool()
  87. def search(query: str) -> list[dict]: ...
  88. def handle(request: dict, session_id: str) -> dict:
  89. with overmind.run(
  90. "support-run", intent=request["question"], conversation_id=session_id
  91. ) as run:
  92. answer = agent(request)
  93. run.deliver(answer) # the final output that gets scored
  94. return answer
  95. ```
  96. `providers="auto"` instruments the LLM SDKs you already use over OpenTelemetry; without a key, tracing is off and nothing breaks. Any OTel exporter can `POST /api/v1/traces` instead, and existing traces in Langfuse, LangSmith, Braintrust or Galileo can be synced through a connector. Open **Observability** to see the trace and its score.
  97. ## Connect your coding agent
  98. Overmind ships an MCP server at `/api/mcp/` with tools, resources and prompts for the platform's agent workflows. Account API keys and OAuth connections can access every active project their account is authorized to use: call `list_projects`, then pass the selected `project_id` on project tools and resource URIs. Membership is checked on every operation. Project API keys remain limited to their configured project. Every tool declares what it costs to run (`free`, `compute`, `llm`, `gpu`) and whether it is destructive. Evaluation matrix rows can be deleted through their scoped lifecycle tool.
  99. The [optional plugin](overmind/README.md#mcp-and-optional-plugins) packages the MCP connection, Overmind branding and the Overmind workflow skills. OAuth connections remain authorized until revoked; access tokens expire after one hour and refresh tokens rotate. Direct MCP connections can use an API key without OAuth or a plugin.
  100. `overmind init --ide <cursor|claude|opencode|codex>` prepares the local configuration and `overmind sync` installs the repository's project API key. To configure an API key by hand:
  101. <details>
  102. <summary><b>Cursor</b> — <code>.cursor/mcp.json</code></summary>
  103. ```json
  104. {
  105. "mcpServers": {
  106. "overmind": {
  107. "url": "https://api.overmindlab.ai/api/mcp/",
  108. "headers": { "X-Api-Key": "ovr_…" }
  109. }
  110. }
  111. }
  112. ```
  113. </details>
  114. <details>
  115. <summary><b>Claude Code</b></summary>
  116. ```bash
  117. claude mcp add --transport http overmind https://api.overmindlab.ai/api/mcp/ --header "X-Api-Key: ovr_…"
  118. ```
  119. </details>
  120. <details>
  121. <summary><b>OpenCode</b> — <code>opencode.json</code></summary>
  122. ```json
  123. {
  124. "$schema": "https://opencode.ai/config.json",
  125. "mcp": {
  126. "overmind": {
  127. "type": "remote",
  128. "url": "https://api.overmindlab.ai/api/mcp/",
  129. "enabled": true,
  130. "headers": { "X-Api-Key": "ovr_…" }
  131. }
  132. }
  133. }
  134. ```
  135. </details>
  136. <details>
  137. <summary><b>Codex</b> — <code>.codex/config.toml</code></summary>
  138. ```toml
  139. [mcp_servers.overmind]
  140. url = "https://api.overmindlab.ai/api/mcp/"
  141. http_headers = { "X-Api-Key" = "ovr_…" }
  142. ```
  143. </details>
  144. <details>
  145. <summary><b>What the tools cover</b></summary>
  146. | Domain | Main tools |
  147. | --------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
  148. | Projects | `list_projects`, `create_api_key` |
  149. | Observability | `inspect_capability_health`, `query_failures`, `query_traces`, `list_signals`, `set_signal_subscription`, `get_job`, `inspect_operation` |
  150. | Datasets | `list_datasets`, `inspect_dataset`, `query_dataset`, `create_dataset_from_traces`, `create_dataset_from_llm_calls`, `explore_dataset`, `create_data_partition`, `inspect_dataset_workbench`, `save_dataset_pipeline`, `run_dataset_pipeline`, `import_dataset_version`, `update_dataset` |
  151. | Evaluations | `check_evaluation_readiness`, `create_eval_set`, `create_eval_matrix_row`, `delete_eval_matrix_row`, `run_evaluation`, `compare_evaluations` |
  152. | Training | `check_finetune_readiness`, `prepare_training_data`, `estimate_finetune`, `start_finetune`, `inspect_training_progress`, `cancel_finetune`, `deploy_finetune`, `retry_deployment`, `set_active_model`, `run_inference`, `get_model_swap_prompt` |
  153. | Decision models | `list_decision_models`, `start_decision_job`, `create_native_evaluation`, `inspect_native_evaluation`, `create_training_experiment`, `measure_decision_performance` |
  154. | Optimiser | `check_optimizer_readiness`, `start_optimizer`, `inspect_optimizer_result` |
  155. | Connectors | `inspect_connectors`, `configure_connector`, `sync_connector` |
  156. | Instrumentation | `get_instrumentation_plan`, `verify_instrumentation` |
  157. | Catalog | `get_model_catalog`, `list_model_workflows` |
  158. `tools/list` returns the tools the connection's permissions allow: a read-only connection omits every write tool.
  159. Prompts such as `investigate-capability`, `finetune-capability` and `ship-model` chain the tools into complete workflows.
  160. </details>
  161. For a self-hosted instance, replace the host with your API URL (`http://localhost:8000` locally). Keys are written to git-ignored files only; `overmind sync` will not write a key into a tracked file.
  162. ______________________________________________________________________
  163. ## Contributing
  164. Open an [issue](https://github.com/overmind-core/overmind/issues/new), or a PR from a feature branch using `.github/PULL_REQUEST_TEMPLATE.md` — `main` is protected and `AGENTS.md` describes how we work. Questions go to the [Discord](https://discord.gg/TPF722ZKuj).
  165. ______________________________________________________________________
  166. ## Telemetry
  167. The SDK and CLI send anonymous usage analytics to PostHog — one `cli.invoked` event per CLI run and `sdk_init` on library use; never prompts, trace contents, keys or dataset contents. Opt out with `OVERMIND_ANALYTICS_ENABLED=false` or `DO_NOT_TRACK=1`; analytics is also off when `CI` is set. Your traces go only to your own project.
  168. ______________________________________________________________________
  169. ## Licence
  170. This repository contains two licences. The platform (`overbae/`, `frontend/`, and everything else outside `overmind/`) is [AGPL-3.0](LICENSE); the root `LICENSE` file is the verbatim AGPL-3.0 text. The SDK, CLI and client libraries under `overmind/` are [MIT](overmind/LICENSE). Copyright (c) 2026 Overmind Ltd. A commercial licence is a paid alternative from Overmind Ltd if you need different terms — [support@overmindlab.ai](mailto:support@overmindlab.ai).
  171. <p align="center">
  172. <img alt="Overmind" src="frontend/src/assets/overmind-eye-copper.svg" width="96">
  173. </p>
  174. <p align="center">
  175. <a href="https://docs.overmindlab.ai">docs.overmindlab.ai</a> · <a href="https://www.overmindlab.ai/?utm_source=github&amp;utm_medium=readme&amp;utm_campaign=always-on&amp;utm_content=readme-footer">overmindlab.ai</a>
  176. </p>

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.