Categories › AI & Agents › Multi-agent, orchestration & workflows
Hosted backend for AI agents: create projects with persistent collections and keys, no signup.
Remote server: your requests and data go to reqres.in. Claude Code: claude mcp add --transport http reqres https://reqres.in/mcp · Claude app: Settings → Connectors → Add custom connector.
The tools this server listed when we connected to it, in its own words.
create_project — Create a real hosted backend (a REST API with a persistent database) in one call, with no account or API key. Use it when the user needs to list_collections — List the collections in the project.create_collection — Create a collection: a new resource (like a table) with its own REST endpoint at /api/collections/{slug}/records. Use one per kind of thing list_records — Read the records stored in a collection (paginated).create_record — Save a record (a JSON object) in a collection so it persists. Pass the fields as data; the API returns it with an id.get_record — Read one record by id.patch_record — Merge the given fields into a record and keep the rest, e.g. {"done": true} to tick off a task. Use this for most edits.create_app_user — Add an end user to the project (for apps where each user sees only their own records). No email is sent. Then call create_app_user_session tcreate_app_user_session — Get a 6-hour session token for an app user without email. Send it as Authorization: Bearer <session_token> to https://reqres.in/app/collectilist_logs — See recent requests to the project (method, path, status, timing), newest first. Use it to debug why an app's calls fail.update_record — Replace all of a record's fields: fields you leave out are removed. To change some fields and keep the rest, use patch_record.delete_record — Delete a record from a collection.Not tested by KitScout yet. When our harness runs it, pros, cons and the tasks we tried appear here, with the date and evidence.