# OpenTrain (npm · @opentrain-ai/mcp)

Hire human data labelers, RLHF annotators, and evaluators from your coding agent.

- Trust score: 64/100 (medium)
- Change this week: +46
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- npm · `@opentrain-ai/mcp`: 64/100 (this document), [markdown](https://verifymcp.io/servers/opentrain-ai-opentrain/opentrain-ai-mcp.md), [page](https://verifymcp.io/servers/opentrain-ai-opentrain/opentrain-ai-mcp)

## Channel facts

- Registry: `npm`
- Package: `@opentrain-ai/mcp`
- Version: `0.2.2`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Supply Chain Security**: 87/100
  - No malware found by supply-chain analysis.
  - Only part of the dependency tree could be resolved (95 of 99), so this covers what we could see, not the whole tree.
  - No install/post-install scripts declared.
  - Only part of the dependency tree could be resolved (95 of 99), so this covers what we could see, not the whole tree.
- **Provenance & Transparency**: 19/100
  - Repository check failed: the declared repository URL returned HTTP 404.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 33 days ago).
  - Security-disclosure policy not yet verified: we couldn't inspect the source repository.
- **Schema Quality & AI Usability**: 75/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 4610 tokens (~112/item across 41 items; 41 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 27/100
  - Stability observed for 8 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 98/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 93% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add opentrain-ai-opentrain -- npx -y @opentrain-ai/mcp
```

### Codex

```bash
codex mcp add opentrain-ai-opentrain -- npx -y @opentrain-ai/mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "opentrain-ai-opentrain": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "@opentrain-ai/mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add opentrain-ai-opentrain --command npx --arg -y --arg @opentrain-ai/mcp
```

### Hermes

```yaml
mcp_servers:
  opentrain-ai-opentrain:
    command: "npx"
    args: ["-y", "@opentrain-ai/mcp"]
```

### Other

```json
{
  "mcpServers": {
    "opentrain-ai-opentrain": {
      "command": "npx",
      "args": [
        "-y",
        "@opentrain-ai/mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-03 (score 64, +1)

No change was recorded against any check on this day. Stability & Change Management went from 23 to 27. That category is still filling its 30-day observation window: 7 days of observed history at the previous scan, 8 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-08-02 (score 63, +40)

- [security improvement] Known CVEs: unverified → partial
- [security improvement] Malware scan: unverified → pass
- [security] Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window).
- [functional improvement] Stability: unverified → 0.23
- [functional improvement] Dependency health: unverified → partial
- [functional improvement] Schema quality: unverified → excellent
- [functional improvement] MCP protocol: unverified → pass

### 2026-08-01 (score 23, +8)

- [security regression] Provenance: unverified → fail
- [security improvement] Install scripts: unverified → pass
- [functional improvement] Maintenance: unverified → pass
- [functional improvement] License: unverified → pass
- [functional] Licence: MIT

### 2026-07-31 (score 15, −24)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-28 (score 39, +21)

- [functional improvement] Tool coverage: unverified → 100
- [functional] First check of Tool coverage: 100
- [functional] First check of Schema quality: unverified
- [functional] First check of Tool coverage: 93
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: fail

### 2026-07-27 (score 18)

First indexed and scored.

## MCP tools (41)

### `opentrain_register_agent` (~149 tokens)

Register a brand-new anonymous OpenTrain agent account. No prior credentials are needed. The new API token is stored in the shared OpenTrain config file (also used by the OpenTrain CLI), so every other opentrain tool works immediately afterwards. Next step: ask the human for their email and call opentrain_claim_account so they can claim the account and add billing.

Input parameters:

- `agentName` (string): Optional display name for the agent identity (e.g. "Claude Code").
- `force` (boolean): Set true to overwrite existing saved OpenTrain credentials with a brand-new agent account.
- `organizationName` (string): Optional name for the employer organization created with the agent account.

Output parameters:

- `apiBaseUrl` (string)
- `claimTokenExpiresAt` (string)
- `credentialsPath` (string)
- `nextSteps` (array)
- `registrationId` (string)
- `scopes` (array)

### `opentrain_claim_account` (~131 tokens)

Send the human owner a claim invite for the agent-registered OpenTrain account. Ask the human for their email first. OpenTrain emails them a verification link and code; once they accept, they own the account and can add billing. Follow up with opentrain_claim_status to check whether the claim completed.

Input parameters:

- `claimToken` (string): Claim token from opentrain_register_agent. Defaults to the token saved in the shared OpenTrain config.
- `email` (string, required): The human owner's email address. OpenTrain emails them a claim invite so they can take over the account and add billing.

Output parameters:

- `apiBaseUrl` (string)
- `emailSent` (boolean)
- `expiresIn` (number)
- `interval` (number)
- `userCode` (string)
- `verificationUri` (string)

### `opentrain_claim_status` (~94 tokens)

Check whether the human has completed the OpenTrain account claim started by opentrain_claim_account. When the claim is complete, the stored API token is upgraded to the claimed account automatically. Respect the poll interval from opentrain_claim_account; a slow_down status means you are polling too fast.

Input parameters:

- `claimToken` (string): Claim token to check. Defaults to the token saved in the shared OpenTrain config.

Output parameters:

- `apiBaseUrl` (string)
- `scopes` (array)
- `status` (string)

### `opentrain_auth_status` (~29 tokens)

Confirm that the configured OpenTrain personal API token can authenticate to the OpenTrain public API.

Output parameters:

- `apiBaseUrl` (string)
- `authenticated` (boolean)
- `ownerOrganizationId` (string|null)
- `scopes` (array)
- `tokenId` (string)
- `tokenLabel` (string|null)
- `userId` (string)

### `opentrain_create_job_draft` (~257 tokens)

Create an unpublished OpenTrain job draft from a full job description/project brief or supported import payload. This is the primary MCP job-drafting workflow: OpenTrain parses the description into its structured job fields server-side. The response includes validation state; each missing field carries an "ask:" question to relay to the human, its type, allowed enum values, and the field name(s) to set via opentrain_update_job_draft_fields. Publish with opentrain_publish_job when the draft is publish ready.

Input parameters:

- `canonicalJob` (object): Optional simple OpenTrain canonical job object. Uses format=opentrain_canonical.
- `externalId` (string): Optional source-system id for audit/idempotency.
- `idempotencyKey` (string): Optional idempotency key. Reuse this key to avoid duplicate drafts on retry.
- `importPayload` (object): Optional supported import payload to send directly to the import-first API, such as schema.org JSON-LD or feed-derived JSON.
- `jobDescription` (string): Full plain-text job description or project brief. This is the primary V1 workflow.
- `title` (string): Optional title to prepend to the job description for plain-text imports.

Output parameters:

- `action` (string)
- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_update_job_draft_fields` (~194 tokens)

Fill in or correct fields on an existing unpublished OpenTrain draft. Use this after opentrain_create_job_draft when the draft is not publish ready: ask the human each missing field's "ask:" question, then patch the answers using the "set:" field names (do not ask the human to author raw JSON). Always start from opentrain_create_job_draft with a description; this tool never publishes directly — use opentrain_publish_job for that.

Input parameters:

- `jobId` (string, required): Existing unpublished draft job id.
- `patch` (object, required): Draft field patch keyed by OpenTrain field names. Use the "set:" field names from the missing-field details returned by opentrain_create_job_draft (e.g. {"experienceLevel":"INTERMEDIATE","paymentType…

Output parameters:

- `action` (string)
- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_publish_job` (~96 tokens)

Publish an OpenTrain draft job live on the marketplace. Runs the same validation + moderation pipeline as the in-app publish flow and is subject to per-account daily publish limits. Requires jobs:write scope and the public_api_job_publishing feature (check the job-drafts capabilities endpoint or a 403 response). Returns the live job URL.

Input parameters:

- `jobId` (string, required): Existing unpublished OpenTrain draft job id to publish live.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_invite_freelancer` (~117 tokens)

Invite a freelancer to a published OpenTrain job, creating a proposal they can respond to. Requires proposals:write scope and the public_api_hiring feature. Agent accounts must be claimed by a human first (403 with a claimUrl otherwise). Idempotent: re-inviting the same freelancer returns the existing proposal with alreadyInvited=true.

Input parameters:

- `freelancerId` (string, required): OpenTrain freelancer user id to invite.
- `jobId` (string, required): Published OpenTrain job id to invite the freelancer to.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_proposal` (~126 tokens)

Read a full OpenTrain proposal/candidate evaluation by proposal id: status, bid, masked candidate, AI-interview score + summary, location/identity verification, Open Label assessment, and contract state. Pass includeInterview: true to also fetch the sanitized AI-interview transcript. Requires proposals:read scope and preserves OpenTrain proposal privacy restrictions.

Input parameters:

- `includeInterview` (boolean): Set true to also fetch the sanitized AI-interview transcript (interviewer/candidate messages) for deeper candidate evaluation.
- `proposalId` (string, required): OpenTrain proposal/job offer id to read.

Output parameters:

- `apiBaseUrl` (string)
- `interview`
- `result` (object)

### `opentrain_get_freelancer_profile` (~109 tokens)

Read a masked OpenTrain freelancer profile by user id or public profile slug for candidate evaluation: title, bio, skills, stats (earned, billed hours, job success), work/label experience, education, reviews, and languages. Requires proposals:read scope. Names are masked to first name + last initial and personal contact details (including personal email) are never returned.

Input parameters:

- `idOrSlug` (string, required): OpenTrain freelancer user id or public profile slug to read.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_hire_proposal` (~269 tokens)

Request hiring a freelancer from an OpenTrain proposal. This does NOT hire or move money: it records a pending approval and returns an approvalUrl a signed-in human must open and confirm in the OpenTrain app (expires in ~72h). The human confirms in the normal hiring interface and may adjust the milestone terms there. Watch opentrain_poll_updates for approval.confirmed or re-check with opentrain_get_approval — the confirmed result includes the new contractId. Requires proposals:write scope and the public_api_hiring feature. Agent accounts must be claimed by a human (403 with claimUrl), and the account needs a payment method or covering credit balance (409 with reason payment_method_required and a billingUrl a human must visit). A 409 with reason not_fit_confirmation_required means retry with confirmNotFitOverride=true if intentional. Re-requesting with identical terms returns the existing pending approval; changed terms supersede it.

Input parameters:

- `confirmNotFitOverride` (boolean): Set true to confirm hiring a proposal previously marked "Not a fit" after a 409 not_fit_confirmation_required response.
- `milestone` (object, required): First escrow milestone for the new contract. Provide name and/or description.
- `proposalId` (string, required): OpenTrain proposal/job offer id to hire from.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_read_messages` (~149 tokens)

Read authorized OpenTrain conversation summaries or messages. Omit conversationId to list summaries; provide conversationId to read messages. Requires messages:read scope and never creates conversations or sends messages.

Input parameters:

- `conversationId` (string): Optional conversation id. Omit it to list authorized conversation summaries instead of messages.
- `cursor` (string): Optional pagination cursor.
- `direction` (string): Message pagination direction when conversationId is provided.
- `filter` (string): Conversation summary filter when conversationId is omitted.
- `limit` (integer): Maximum records to return. Defaults to the OpenTrain API route default.
- `unreadOnly` (boolean): When listing conversations, return only conversations with unread messages.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_send_message` (~126 tokens)

Send a plain-text message into an existing OpenTrain conversation the token owner participates in. Runs the same membership, rate-limit, and content-policy checks as the in-app messaging flow. Requires messages:write scope and the public_api_messaging_writes feature. Agent accounts must be claimed by a human first (403 with a claimUrl otherwise). This tool never creates conversations; conversations come from proposals, invites, and hires.

Input parameters:

- `content` (string, required): Plain-text message content to send.
- `conversationId` (string, required): Existing OpenTrain conversation id the token owner participates in.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_pending_payments` (~48 tokens)

Read pending OpenTrain payment and milestone state visible to the token owner. Requires payments:read scope and never releases funds, refunds charges, creates payouts, or changes payment settings.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_jobs` (~125 tokens)

List the token owner's OpenTrain jobs with status, publish state, proposal counts, and live URLs. Requires jobs:read scope. Use this to find job ids for opentrain_list_proposals or opentrain_publish_job.

Input parameters:

- `cursor` (string): Optional pagination cursor.
- `limit` (integer): Maximum jobs to return. Defaults to 25.
- `status` (string): Optional job status filter: DRAFT, OPEN, ONGOING, COMPLETED, ARCHIVED, or PENDING_APPROVAL. Omit to list all non-deleted jobs.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_proposals` (~147 tokens)

List proposals/candidates for an OpenTrain job with statuses, bids, AI interview scores, and resume match scores — the primary tool for reviewing and ranking who to hire. Requires proposals:read scope. Use opentrain_get_proposal for one candidate in depth, then opentrain_hire_proposal to hire.

Input parameters:

- `cursor` (string): Optional pagination cursor.
- `jobId` (string, required): OpenTrain job id to list proposals for.
- `limit` (integer): Maximum proposals to return. Defaults to 25.
- `status` (string): Optional proposal status filter such as UNREVIEWED, SHORTLISTED, HIRED, or DECLINED.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_team` (~49 tokens)

Read the token owner's OpenTrain employer team: organization, members, and pending invites. Team members share jobs and the team inbox. Requires team:read scope and the public_api_team feature.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_invite_team_member` (~121 tokens)

Invite a human to the OpenTrain employer team by email, giving them shared access to jobs and the team inbox once they accept. Sends the same invitation email as the in-app flow. Requires team:write scope and the public_api_team feature. Agent accounts must be claimed by a human first (403 with a claimUrl otherwise). Idempotent-ish: inviting an existing member returns already_member; users with existing accounts are added directly (member_added).

Input parameters:

- `email` (string, required): Email address of the human to invite to the employer team.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_start_proposal_conversation` (~106 tokens)

Start (or fetch) the pre-hire direct-message thread for a proposal so you can message the candidate before hiring. Idempotent get-or-create — safe to call repeatedly; returns the conversationId to use with opentrain_send_message. Employer side only. Requires messages:write scope, the public_api_messaging_writes feature, and a claimed account.

Input parameters:

- `proposalId` (string, required): Proposal id to open the pre-hire conversation thread for.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_poll_updates` (~139 tokens)

Poll the OpenTrain delta feed of account events (new proposals, proposal status changes, new messages, contracts, milestone changes, pending payments) in one cheap call instead of re-reading every resource. Pass the nextCursor from the previous poll to receive only newer events; payloads carry IDs only — fetch details with the matching read tool. Event visibility follows the token scopes (proposals:read, messages:read, payments:read).

Input parameters:

- `cursor` (string): Last event id already processed (nextCursor from the previous poll). Omit on the first poll.
- `limit` (integer): Max events to return (default 50).

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_search_jobs` (~115 tokens)

Search the public OpenTrain job marketplace (all live jobs, not just the token owner's). Filter by free text, category, language, country, or pay type. Use opentrain_list_jobs for the token owner's own jobs instead.

Input parameters:

- `category` (string)
- `country` (string): ISO country code.
- `cursor` (string)
- `language` (string)
- `limit` (integer)
- `payType` (string)
- `q` (string): Free-text search query.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_capabilities` (~64 tokens)

Discover which OpenTrain public API features are enabled for this token (job drafting, publishing, hiring, messaging writes, team) plus accepted job-draft fields and enums. Call this when a write tool returns a feature-disabled error or before planning a multi-step flow.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_tokens` (~47 tokens)

List the API tokens on the OpenTrain account (labels, scopes, expiry — never secrets). Requires tokens:read scope. Use opentrain_revoke_token to revoke one.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_revoke_token` (~65 tokens)

Revoke an OpenTrain API token by id (irreversible — the token stops working immediately). Requires tokens:write scope. Revoking the token this server is using will break subsequent calls.

Input parameters:

- `tokenId` (string, required): Id of the API token to revoke.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_contracts` (~107 tokens)

List contracts (hired freelancers) on the OpenTrain account, optionally filtered by jobId and/or status (active|ended). Each contract includes milestones, the post-hire job DM conversationId, and the freelancer identity (unmasked post-hire). Requires payments:read scope.

Input parameters:

- `jobId` (string): Optional OpenTrain job id to list contracts for.
- `status` (string): Optional contract status filter: active or ended. Omit to list both.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_contract` (~63 tokens)

Read one OpenTrain contract in detail: status, milestones with funding/approval state, freelancer identity, and the post-hire job DM conversationId for messaging. Requires payments:read scope.

Input parameters:

- `contractId` (string, required): OpenTrain contract id to read.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_create_milestone` (~161 tokens)

Create a new unfunded milestone on an existing OpenTrain contract. No money moves at creation — use opentrain_request_milestone_funding afterwards to start the human-confirmed funding step. Requires payments:write scope and a claimed account.

Input parameters:

- `amountUsd` (number): Optional milestone amount in USD. No money moves at creation.
- `contractId` (string, required): Contract id to add the milestone to.
- `description` (string, required): Description of the work to be delivered in this milestone.
- `dueDate` (string): Optional due date (ISO 8601).
- `name` (string): Optional short milestone name.
- `volume` (number): Optional unit volume (e.g. label count) for per-unit milestones.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_request_milestone_funding` (~111 tokens)

Request escrow funding for a milestone. This does NOT move money: it returns a pending approval with an approvalUrl that a signed-in human must open and confirm in the OpenTrain app (expires in ~72h). Watch opentrain_poll_updates for an approval.confirmed event, or re-check with opentrain_get_approval. Requires payments:write scope, a claimed account, and the payments-write feature flag.

Input parameters:

- `milestoneId` (string, required): OpenTrain milestone id.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_request_milestone_approval` (~115 tokens)

Request approval (payment release) of a funded milestone. This does NOT release money: it returns a pending approval with an approvalUrl that a signed-in human must open and confirm in the OpenTrain app (expires in ~72h). Watch opentrain_poll_updates for an approval.confirmed event, or re-check with opentrain_get_approval. Requires payments:write scope, a claimed account, and the payments-write feature flag.

Input parameters:

- `milestoneId` (string, required): OpenTrain milestone id.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_approval` (~75 tokens)

Check the status of a pending human co-sign approval (milestone funding, milestone release, or contract end): pending, confirmed, declined, or expired, plus the execution result once confirmed. Requires payments:read scope.

Input parameters:

- `approvalId` (string, required): Approval id returned by a fund/approve/end-contract request.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_end_contract` (~112 tokens)

End an OpenTrain contract. If the contract has no funded milestones it ends immediately. If funded escrow is at stake, the call instead returns a pending approval with an approvalUrl a signed-in human must confirm (no funds move until then) — watch opentrain_poll_updates for approval.confirmed or re-check with opentrain_get_approval. Requires payments:write scope, a claimed account, and the payments-write feature flag.

Input parameters:

- `contractId` (string, required): OpenTrain contract id to end.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_close_job` (~82 tokens)

Close (archive) a published OpenTrain job so it stops accepting proposals and leaves public listings. Existing contracts on the job are unaffected. Idempotent — closing an already-archived job reports alreadyClosed. Requires jobs:write scope and the job-publishing feature.

Input parameters:

- `jobId` (string, required): Published OpenTrain job id to close (archive).

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_update_published_job` (~155 tokens)

Update fields on a live published (OPEN) OpenTrain job. The revised listing is re-checked by moderation inline; if it is blocked, the job is automatically unpublished back to draft and the response explains what to fix. For unpublished drafts use opentrain_update_job_draft_fields instead. Requires jobs:write scope and the job-publishing feature.

Input parameters:

- `jobId` (string, required): Published (OPEN) OpenTrain job id to update.
- `patch` (object, required): Job field patch keyed by OpenTrain field names, same shape as opentrain_update_job_draft_fields (e.g. {"jobTitle":"...","pricePerHour":14}). The revised listing is re-checked by moderation.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_credits` (~85 tokens)

Read the OpenTrain credit balance for the account: available cents, reserved (escrow-held) cents, currency, and the 10 most recent ledger entries. Credits fund hires and milestone escrow without a card-present step; when the balance covers a money move, the existing human co-sign approval flow executes against credits. Requires payments:read scope and the credits feature.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_credit_ledger` (~110 tokens)

Page through the OpenTrain credit ledger: top-ups, escrow holds, hold releases, captures, refunds, and adjustments, newest first. Each entry links the related top-up, proposal, contract, or milestone ids. Pass the returned nextCursor to fetch older entries. Requires payments:read scope and the credits feature.

Input parameters:

- `cursor` (string): Opaque pagination cursor from a previous nextCursor value.
- `limit` (integer): Page size 1-100 (default 50).

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_create_credit_top_up` (~132 tokens)

Start a credit top-up on the OpenTrain account. No money moves from this call: it returns a Stripe Checkout checkoutUrl that a signed-in human must open and pay (the link expires in ~24h). Once paid, the balance updates automatically — confirm with opentrain_get_credit_top_up or opentrain_get_credits. Amount must be between $10 and $10,000. Requires payments:write scope, a claimed account, and the credits feature.

Input parameters:

- `amountUsd` (number, required): Top-up amount in US dollars (min $10, max $10,000).

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_credit_top_up` (~69 tokens)

Check the status of a credit top-up: pending (awaiting human payment), completed (credits added), canceled, or expired. Requires payments:read scope and the credits feature.

Input parameters:

- `topUpId` (string, required): Credit top-up id returned when the top-up was created.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_create_webhook` (~157 tokens)

Subscribe a URL to OpenTrain platform events (push instead of polling opentrain_poll_updates). Deliveries are signed with HMAC-SHA256 via the X-OpenTrain-Signature header. IMPORTANT: the signing secret is returned ONCE in this response — store it immediately; it cannot be retrieved later. Requires webhooks:manage scope, the webhooks feature, and the read scope for each subscribed event type. The subscription only receives events created after it.

Input parameters:

- `eventTypes` (array, required): Event types to subscribe to (e.g. ["proposal.received","message.received"]). Each requires the matching read scope on the token.
- `url` (string, required): HTTPS endpoint URL that will receive signed event deliveries.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_list_webhooks` (~55 tokens)

List webhook subscriptions on the OpenTrain account, including status (ACTIVE or DISABLED after sustained delivery failure) and subscribed event types. Secrets are never included. Requires webhooks:manage scope and the webhooks feature.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_get_webhook` (~76 tokens)

Read one webhook subscription: URL, event types, status (ACTIVE or DISABLED), and failure/disable details. The signing secret is never returned — if lost, delete and re-create the subscription. Requires webhooks:manage scope and the webhooks feature.

Input parameters:

- `webhookId` (string, required): Webhook subscription id.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

### `opentrain_delete_webhook` (~73 tokens)

Delete a webhook subscription and stop its deliveries. This is also how a DISABLED subscription is resumed: delete it and create a new one (which mints a new secret). Requires webhooks:manage scope and the webhooks feature.

Input parameters:

- `webhookId` (string, required): Webhook subscription id to delete.

Output parameters:

- `apiBaseUrl` (string)
- `result` (object)

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/opentrain-ai-opentrain/opentrain-ai-mcp#diagnostics

## Score history

- 2026-08-03: 64
- 2026-08-02: 63
- 2026-08-01: 23
- 2026-07-31: 15
- 2026-07-30: 39
- 2026-07-28: 39
- 2026-07-27: 18

## Links

- npm package: https://www.npmjs.com/package/@opentrain-ai/mcp
- Socket report: https://socket.dev/npm/package/@opentrain-ai/mcp
- Website: https://www.opentrain.ai/docs/developers/overview
- Changelog RSS feed: https://verifymcp.io/servers/opentrain-ai-opentrain/opentrain-ai-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/opentrain-ai-opentrain/opentrain-ai-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/opentrain-ai-opentrain/opentrain-ai-mcp