Inspect Depot builds, CI runs, job logs and Actions runners; retry or cancel CI runs. Hosted by usefulapi — connect from Claude, ChatGPT, Cursor, or any MCP client.
Team and Enterprise: an Owner adds the connector for the organization first.
Run this command, then run /mcp in Claude Code to log in:
Add to Cursor or add this to ~/.cursor/mcp.json:
{
"mcpServers": {
"depot": {
"url": "https://depot.usefulapi.io/mcp"
}
}
}Add to VS Code or add this to .vscode/mcp.json:
{
"servers": {
"depot": {
"type": "http",
"url": "https://depot.usefulapi.io/mcp"
}
}
}No Add custom MCP server option? It depends on your account or workspace settings: see OpenAI's guide.
Run these commands. The second one opens the login page:
Other MCP clients (Windsurf, Cline, Zed and more): add the URL as a remote MCP server (Streamable HTTP). The client then opens the login page in your browser. Do not add an Authorization header: see Before you connect.
The login page asks for your Depot Organization API token.
Create one in the Depot dashboard → Organization Settings → API Tokens. Depot derives the organization from the token. Your token is used only to call the API on your behalf and to derive a private account id; it is never shown to anyone.
Add only the URL. Do not add an Authorization header or an API key to the client config. The server signs you in with OAuth: the login page asks for your Depot Organization API token. If the config has such a header, remove it: some clients then send that header instead of the login token, and every call fails with 401.
Turn on the Depot connector, then ask in plain words. The AI picks the tools.
| Tool | Type | What it does |
|---|---|---|
depot_list_projects | read | List projects List the organization's Depot projects. A project is an isolated container-build cache (region, hardware, cache policy). Use the returned projectId with the build and registry tools. Depot: ProjectService/ListProjects. |
depot_get_project | read | Get a project Get one Depot project — name, region, hardware size, cache policy, build timeout and autoscaling. Depot: ProjectService/GetProject. |
depot_list_builds | read | List builds for a project List container builds in a project, newest first: status (STATUS_RUNNING / FAILED / SUCCESS / ERROR / CANCELED), timestamps, build duration, time saved by cache, and cached vs total steps. Depot: core BuildService/ListBuilds. |
depot_get_build | read | Get a build Get one container build's status, timestamps, duration, time saved and cached/total step counts. Depot: core BuildService/GetBuild. |
depot_get_build_steps | read | List a build's steps List the BuildKit steps of a container build: name, digest, start/end time, cache state (CACHED / UNCACHED), any error, and whether it has logs. Find the failing step here, then pass its digest to depot_get_build_step_logs. Depot: build.v1 BuildService/GetBuildSteps. |
depot_get_build_step_logs | read | Get a build step's logs Get the log lines (message + timestamp) of one container-build step, identified by its digest from depot_get_build_steps. Depot: build.v1 BuildService/GetBuildStepLogs. |
depot_list_registry_images | read | List registry images List the images stored in a project's Depot Registry: tag, digest, pushedAt and sizeBytes. Depot: build.v1 RegistryService/ListImages. |
depot_get_usage | read | Get organization usage Get the organization's Depot usage over a window: container builds per project (build count, minutes saved, minutes billed), GitHub Actions jobs per repo/workflow/runner (minutes elapsed and billed), storage, and agent sandboxes. Depot: UsageService/GetUsage. |
depot_ci_list_runs | read | List Depot CI runs List Depot CI runs, newest first. Filter by status (default: running + queued — pass ["failed"] to find failures), repo, commit SHA prefix, trigger event, or pull request number (requires repo). Depot: CIService/ListRuns. |
depot_ci_get_run_status | read | Get a CI run's status tree Get a Depot CI run's status with its nested workflows, jobs and attempts (ids, keys, statuses, error messages). This is how to find the workflowId / jobId / attemptId for the log, diagnosis and retry tools. Depot: CIService/GetRunStatus. |
depot_ci_list_workflows | read | List Depot CI workflows List Depot CI workflows across runs, newest first, with per-status job counts. Filter by name/path text, repo, status, trigger, SHA prefix or PR number. Depot: CIService/ListWorkflows. |
depot_ci_get_job | read | Get a CI job Get one Depot CI job: status, conclusion, error message, timestamps, runner config (labels, image, CPUs, memory), matrix values, parent run/workflow context, dependencies and attempt history. Depot: CIService/GetJob. |
depot_ci_get_job_logs | read | Get a CI job's logs Get persisted log lines for a Depot CI job attempt, oldest first (step key/name, timestamp, line number, body). Pass job_id to read the job's latest attempt, or attempt_id for a specific one — exactly one. Pass nextPageToken back as page_token for more lines. Depot: CIService/GetJobAttemptLogs. |
depot_ci_get_failure_diagnosis | read | Diagnose a CI failure Get Depot's failure diagnosis for a CI run, workflow, job or attempt. For runs and workflows, identical failures are grouped by error fingerprint; for a job or attempt it is a focused diagnosis with evidence. Empty when no failure evidence is found. The best first call when a run failed. Depot: CIService/GetFailureDiagnosis. |
depot_gha_list_jobs | read | List GitHub Actions jobs on Depot runners List GitHub Actions jobs that ran on Depot-managed runners, newest first: status, conclusion, runner label, requested CPU/memory, duration, billable time and GitHub URL. Defaults to the last 30 days (max window 90 days). Depot: GithubActionsService/ListGithubActionsJobs. |
depot_gha_search_logs | read | Search GitHub Actions logs Search GitHub Actions log lines from Depot runners, newest match first (job, workflow, step, timestamp, body, lineId). Defaults to the last hour (max window 30 days). Filter by repository, workflow, runner label, action or compute id. Depot: GithubActionsService/SearchGithubActionsLogs. |
depot_gha_get_analytics | read | Get GitHub Actions analytics Get time-bucketed GitHub Actions analytics for Depot runners: job counts (successful/failed/cancelled), failure rate, elapsed and billable seconds, average duration — as totals plus hourly or daily buckets. Defaults to the last 30 days (max 90). Depot: GithubActionsService/GetGithubActionsAnalytics. |
depot_gha_list_recommendations | read | List runner-size recommendations List Depot's runner-size recommendations from job CPU and memory metrics: SIZE_UP when average CPU or memory reaches 90%, SIZE_DOWN when both averages stay ≤30% and peaks ≤70%, with current vs suggested size and per-minute price. Defaults to the last 30 days. Depot: GithubActionsService/ListGithubActionsRecommendations. |
depot_cache_list_entries | read | List cache entries List Depot Cache entries of one type, optionally by key prefix: key, size in bytes, name and creation time (Depot returns no last-access time and no repository). Retention and size limits are set only in the dashboard (Organization Settings), not through the API. Depot: CacheService/ListEntries on cache.depot.dev. |
depot_cache_delete_entries | write | Delete cache entries WRITE — deletes cached build data; a later build rebuilds it. Delete Depot Cache entries by exact key. Without confirm it only previews: it shows which keys exist, with their size, and deletes nothing. Show the preview to the user, then call again with confirm: true. Depot: CacheService/DeleteEntry on cache.depot.dev. |
depot_create_project | write | Create a project WRITE. Create a Depot container-build project (an isolated build cache). Idempotent by name: if an active project with the same name exists, that project is returned. Defaults: cache 50 GB / 14 days, hardware HARDWARE_16X32 on AWS builders. Depot: ProjectService/CreateProject. |
depot_ci_retry_job | write | Retry a CI job WRITE — starts compute. Retry one failed or cancelled Depot CI job, plus any skipped jobs that depend on it. Fails if a dependent job already started; use depot_ci_retry_failed_jobs then. Depot: CIService/RetryJob. |
depot_ci_retry_failed_jobs | write | Retry a workflow's failed jobs WRITE — starts compute. Retry only the failed and cancelled jobs in a Depot CI workflow, plus skipped jobs that depend on them. Jobs that succeeded are left alone. Depot: CIService/RetryFailedJobs. |
depot_ci_cancel_run | write | Cancel a CI run WRITE — stops work in progress. Cancel a queued or running Depot CI run and all of its unfinished workflows, jobs and attempts. Cancelled jobs can be retried afterwards with depot_ci_retry_failed_jobs. Depot: CIService/CancelRun. |
depot_usage_status | Usage status (free-tier meter) Report the caller's current usage: the plan, calls used this month, the limit of the plan (the free monthly cap or the Pro fair use limit), remaining, and whether the limit is reached. Read-only; does not count against the meter. | |
depot_request_feature | Request a missing feature Tell usefulapi that the user needs something the Depot tools cannot do yet (a missing tool, field or filter). Call it only when the user asks for something none of the tools can do, and tell the user that you send the request. Do not call it for errors or for normal requests. Do not include personal data, credentials or customer records. Stored with the product name and the client type (for example Claude or Cursor), without the user's account. At most 5 requests per day. Does not count against the meter. | |
depot_upgrade | Upgrade to Pro (unlimited*) Get Pro: unlimited tool calls (fair use limit: 10,000 a month) on Depot (the free tier caps monthly usage). Plans: "monthly" — $9/month; "yearly" — $90/year (2 months free). Cancel any time: Pro then runs to the end of the paid period, so it also suits a one-time job. *A fair use limit applies. For heavy use (near Pro's limit): "plus" or "plus_yearly" — Pro+, $19/month or $190/year, unlimited tool calls (fair use limit: 100,000 a month); an existing Pro subscription changes to Pro+ in place (needs confirm: true; the price difference for the rest of the period is charged at once). Otherwise returns a short link to open in your browser (it opens Stripe Checkout); after payment the account upgrades automatically. Does not count against the meter. | |
depot_cancel_subscription | Cancel the Pro subscription Cancel the caller's Depot Pro subscription at the end of the paid period (no refund for the current period; Pro continues until then, then the free tier applies). Requires confirm: true. Run depot_upgrade later to undo the cancel before the period ends. Does not count against the meter. |
| Plan | Price | Limit |
|---|---|---|
| Free | $0 | 50 tool calls / month |
| Proper user | $9/mo · $90/yr | Unlimited* |
*Pro fair use limit: 10,000 tool calls a month on this server. For more, write to [email protected].
Pro covers this Depot server only. Subscribe with depot_upgrade (it returns a Stripe Checkout link). Cancel any time with depot_cancel_subscription: Pro continues to the end of the paid period, with no refund for the current period, and running depot_upgrade before then undoes the cancel. Or write to [email protected].
No. usefulapi is an independent service. It is not affiliated with or endorsed by Depot. The server calls the Depot API with your own Depot access, so it sees only the data that your account can see.
Yes. In Claude, add this server's URL as a custom connector. In ChatGPT, add it as a custom MCP server in Plugins. Then log in once. Cursor, VS Code, Claude Code and Codex work too. See the setup steps.
The login page asks for your Depot Organization API token. See Before you connect for where to find it.
No. Add only the URL. Do not add an Authorization header or an API key to the client config. The server signs you in with OAuth: the login page asks for your Depot Organization API token. If the config has such a header, remove it: some clients then send that header instead of the login token, and every call fails with 401.
The login stores them encrypted in the authorization grant of your connection. The server uses them to call the Depot API for you and to derive a private account id for usage metering. usefulapi does not show them on any page or in any reply. To stop all access, remove the connector and change or delete these credentials in Depot.
Yes, if you approve it. 5 of the 24 Depot tools can create or change data. The other 19 are marked read-only. Most MCP clients ask you to approve a tool call before it runs.
The Free plan gives 50 tool calls / month. Pro costs $9/mo or $90/yr, with unlimited tool calls (fair use limit: 10,000 tool calls a month), for this Depot server only. Cancel any time: Pro then runs to the end of the paid period. Run depot_usage_status to see how many calls you used.
Ask the AI to run depot_upgrade: it returns a Stripe Checkout link. To cancel, run depot_cancel_subscription. Pro continues to the end of the paid period.
Tell the AI what you wanted to do. It can send the request with depot_request_feature. We store the text with the server name, the type of AI client and the tool you tried, not with your account, and read every request when we plan new tools. You can send up to 5 requests per day.
Any client that supports remote MCP servers (Streamable HTTP) with OAuth login: Claude (web and desktop), Claude Code, ChatGPT, Codex, Cursor, VS Code, Windsurf and others.
This is a Model Context Protocol endpoint — meant to be connected from an AI client, not opened in a browser. An invalid_token response at the URL is the auth gate working as designed; clients authenticate automatically.