# Adaption MCP MCP Server

> **Adaption MCP MCP Server** is a hosted, multitenant Model Context Protocol (MCP) server run by **MewCP** (https://mewcp.com), giving AI agents managed access to Adaption MCP.
>
> MewCP takes care of all MCP infrastructure for you — credential storage, OAuth flows,
> token refresh, and production-grade auto-scaling — so your AI agents can connect to
> third-party services and run freely without you managing any MCP server yourself.
>
> Server page:  https://mewcp.com/mcp/adaptionlabs-mcp
> MewCP docs:   https://docs.mewcp.com
> All servers:  https://mewcp.com/servers (JSON: https://mewcp.com/api/servers)
> About MewCP:  https://mewcp.com/llms.txt

---

## About

Manage machine learning datasets and fine-tuning workflows through Adaption’s MCP server. Import data from Hugging Face, Kaggle, or Google Sheets; adapt, augment, translate, localize, and generate synthetic datasets, then export results and launch or monitor AutoScientist training runs with model and hyperparameter recommendations

---

## Details

- Server ID: `adaptionlabs-mcp`
- Version: 1.0.0
- Tools: 26
- Authentication: OAuth, managed by MewCP
- Transport: http

---

## Access

The gateway URL below serves your **entire MewCP toolset**, not this server on its
own. Pasting a config snippet is not enough — Adaption MCP has to be in the toolset
first. In order:

1. Add Adaption MCP to your toolset at https://mewcp.com/mcp/adaptionlabs-mcp
2. Connect your Adaption MCP account (OAuth); MewCP stores the credential and attaches it to each call
3. Add the gateway URL to your client (see Connect below). It signs in with OAuth
   the first time it connects, so there is no key to copy.

Once connected, an agent does **not** see this server's tools as top-level tools.
It sees four meta-tools and reaches everything through them:

- `search(query)` — find tools by keyword across the toolset
- `get_schema(tools)` — full description and arguments for the tools you picked
- `manage_connections(requests?)` — connection status and account aliases, plus
  links the user opens to add an app or connect an account
- `call_tool(app_maskedId, tool_name, args, alias?)` — execute

If Adaption MCP isn't in the toolset or connected yet, `manage_connections` with
app_maskedId `adaptionlabs-mcp` returns a link that adds it and connects an account.

So the list below is what `search` can return for this server, not a set of
callable tool names on their own.

---

## Tools (26)

Descriptions are truncated to 160 characters; call `get_schema`
for the full text and the argument schema. Where a tool is annotated, its type
is shown — treat `destructive` as irreversible.

- `list_datasets` _(read)_ — List datasets visible to the authenticated Adaption organization. Returns concise status and pagination fields.
- `get_dataset` _(read)_ — Get a concise public view of one dataset, including processing and column information.
- `get_dataset_status` _(read)_ — Poll dataset processing status and progress. Use after an asynchronous dataset operation.
- `get_dataset_evaluation` _(read)_ — Get quality evaluation status and results for one dataset.
- `get_dataset_export_links` _(read)_ — Poll saved external export URLs for a dataset after an export operation.
- `get_dataset_retrieval_index` _(read)_ — Poll the retrieval index of a dataset after index_dataset. Once status is succeeded, search the dataset by passing its id in file_search.dataset_ids on…
- `import_dataset` _(write)_ — Import a dataset from a HuggingFace, Kaggle, Google Sheets, or public S3 or Cloud Storage object URL. An object URL is one https .csv, .json, .jsonl, or…
- `run_dataset_adaptation` _(write)_ — Estimate or launch adaptation for an imported dataset. Poll dataset status after launch. translate and localize add translated or localized rows, and expand…
- `expand_dataset` _(write)_ — Estimate or launch creation of a new dataset containing the source rows plus curated domain/general rows, or generate similar rows based on the seed dataset.
- `augment_dataset` _(write)_ — Deprecated: use expand_dataset, which takes the same arguments and behaves identically. Estimate or launch creation of a new dataset containing the source rows…
- `translate_dataset` _(write)_ — Estimate or launch creation of a new dataset with sampled rows translated into target languages.
- `localize_dataset` _(write)_ — Estimate or launch creation of a new dataset with sampled rows localized for country/language pairs.
- `combine_datasets` _(write)_ — Combine compatible ready datasets into one new dataset. This is idempotent when the same idempotency key is reused.
- `index_dataset` _(write)_ — Build a retrieval index so a dataset can be searched. Pick the text columns to search (for most uploaded datasets, the file’s column headers); they are…
- `list_training_models` _(read)_ — List base models currently available for AutoScientist training.
- `list_autoscientist_runs` _(read)_ — List AutoScientist runs, optionally filtered to one dataset.
- `get_autoscientist_run` _(read)_ — Poll one AutoScientist run, including iteration progress and any public failure message.
- `recommend_autoscientist_hyperparameters` _(read)_ — Resolve a suitable model and hyperparameters for a ready dataset without launching a run.
- `create_autoscientist_run` _(write)_ — Launch a new asynchronous AutoScientist training run on a ready dataset. This spends credits; poll get_autoscientist_run afterward.
- `list_invent_domains` _(read)_ — List valid domain and qualified subdomain codes for dataset generation.
- `generate_dataset` _(write)_ — Estimate or launch generation of a new invented dataset. Launches consume credits.
- `list_stacks` _(read)_ — List your stacks in the authenticated organization, newest first. Pass a stack model_id as model to the Responses API to call it.
- `get_stack` _(read)_ — Get one stack: its model_id, status, model, tools and feature settings.
- `create_stack` _(write)_ — Create a stack: a hosted model with tools and feature settings, called through its model_id.
- `update_stack` _(write)_ — Change a stack. Only the fields sent change; a model or tool the caller cannot use is refused before anything is written.
- `delete_stack` _(destructive)_ — Delete a stack for good. Requests to its model_id fail afterwards.

---

## Connect

- URL: https://gateway.mewcp.com/personal/mcp
- Transport: Streamable HTTP (remote). Nothing to install.
- Sign-in: OAuth. On first connect the client opens a browser window to sign in
  to MewCP and approve access. No API key.
- Headless clients (SDKs, servers without a browser) send an API key instead:
  `Authorization: Bearer <API_KEY>`, key from https://mewcp.com/dashboard/api-keys

Agents setting themselves up: follow https://mewcp.com/SKILL.md

Step-by-step guide per client:

- Claude Desktop: https://mewcp.com/mcp/adaptionlabs-mcp/connect/claude
- ChatGPT: https://mewcp.com/mcp/adaptionlabs-mcp/connect/chatgpt
- Grok: https://mewcp.com/mcp/adaptionlabs-mcp/connect/grok
- DeepSeek Harness: https://mewcp.com/mcp/adaptionlabs-mcp/connect/dsh
- Mistral Vibe: https://mewcp.com/mcp/adaptionlabs-mcp/connect/mistral-vibe
- VS Code: https://mewcp.com/mcp/adaptionlabs-mcp/connect/vscode
- Cursor: https://mewcp.com/mcp/adaptionlabs-mcp/connect/cursor
- Codex: https://mewcp.com/mcp/adaptionlabs-mcp/connect/codex
- Claude Code: https://mewcp.com/mcp/adaptionlabs-mcp/connect/claude-code
- OpenCode: https://mewcp.com/mcp/adaptionlabs-mcp/connect/opencode
- OpenClaw: https://mewcp.com/mcp/adaptionlabs-mcp/connect/openclaw
- Antigravity: https://mewcp.com/mcp/adaptionlabs-mcp/connect/antigravity
- Hermes: https://mewcp.com/mcp/adaptionlabs-mcp/connect/hermes
- Kimi Code: https://mewcp.com/mcp/adaptionlabs-mcp/connect/kimi-code
- Grok Build: https://mewcp.com/mcp/adaptionlabs-mcp/connect/grok-build
- Devin Desktop: https://mewcp.com/mcp/adaptionlabs-mcp/connect/devin-desktop
- Zed: https://mewcp.com/mcp/adaptionlabs-mcp/connect/zed
- Muse Code: https://mewcp.com/mcp/adaptionlabs-mcp/connect/muse-code
- Cline: https://mewcp.com/mcp/adaptionlabs-mcp/connect/cline
- Kilo Code: https://mewcp.com/mcp/adaptionlabs-mcp/connect/kilo-code
- Kiro: https://mewcp.com/mcp/adaptionlabs-mcp/connect/kiro
- Junie: https://mewcp.com/mcp/adaptionlabs-mcp/connect/junie
- Amp: https://mewcp.com/mcp/adaptionlabs-mcp/connect/amp
- Qwen Code: https://mewcp.com/mcp/adaptionlabs-mcp/connect/qwen-code
- Goose: https://mewcp.com/mcp/adaptionlabs-mcp/connect/goose
- Python: https://mewcp.com/mcp/adaptionlabs-mcp/connect/python
- TypeScript: https://mewcp.com/mcp/adaptionlabs-mcp/connect/typescript