Skip to content
Scalekit Docs
Talk to an Engineer Dashboard

Connect AI agents to the Devin MCP server

Vendor MCP
Open markdown

The Devin MCP connector routes your AI agent's tool calls to Devin's own MCP server through Scalekit. Each user connects their own Devin access token once, and Scalekit sends it with every call, so your agent never handles credentials. It comes with 30 tools.

Tools
30
What they doRead · write · destructive
1 · 29 · 01 read29 write0 destructive
Users sign in with

Setup

  1. Install the SDK

    Terminal window
    npm install @scalekit-sdk/node dotenv
  2. Set your credentials

    Add your Scalekit credentials to your .env file. Find values in app.scalekit.com > Developers > API Credentials.

    .env
    SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
    SCALEKIT_CLIENT_ID=<your-client-id>
    SCALEKIT_CLIENT_SECRET=<your-client-secret>
  3. Create the Devin MCP connection

    In AgentKit > Connections, create a Devin MCP connection. The name you give it is the connection_name your code passes. See Configure connections.

    Console steps with screenshots

    Register your Devin API key with Scalekit so it stores it securely and injects it into every request. Devin uses API key authentication.

    1. Get your Devin API key

      • Sign in to Devin and go to Settings → Devin API.
      • Click the API keys (legacy) tab.
      • Under Personal API key, click View key to reveal your key, or expand the dropdown to regenerate it.

      Devin Settings → Devin API page showing the Personal API key with a View key button

    2. Create a connection in Scalekit

      • In Scalekit dashboard, go to AgentKit → Connections → Create Connection. Find Devin MCP and click Create.
      • Note the Connection name — you will use this as connection_name in your code (e.g., devinmcp).
      • Click Save.
    3. Add a connected account

      Connected accounts link a specific user identifier in your system to their Devin API key. Add one in the dashboard to test. In production, each user adds their own through the authorization link: they enter their credentials on the page it opens.

      In the dashboard, to test

      • Open the connection you created and click the Connected Accounts tab → Add account.
      • Fill in:
        • Your User’s ID — a unique identifier for this user in your system (e.g., user_123)
        • API Key — the Devin API key from step 1
      • Click Create Account.

      From your backend, if your app already has the credentials

      For example, when users enter them on a settings page in your app:

      import { Scalekit, ConnectorStatus } from '@scalekit-sdk/node';
      const scalekit = new Scalekit(
      process.env.SCALEKIT_ENVIRONMENT_URL,
      process.env.SCALEKIT_CLIENT_ID,
      process.env.SCALEKIT_CLIENT_SECRET,
      );
      // Never hard-code credentials — read from secure storage or user input
      const devinApiKey = getUserDevinApiKey(); // retrieve from your secure store
      const authorizationDetails = {
      details: {
      case: 'staticAuth',
      value: {
      details: {
      token: devinApiKey,
      },
      },
      },
      };
      let { connectedAccount } = await scalekit.actions.upsertConnectedAccount({
      connectionName: 'devinmcp',
      identifier: 'user_123',
      authorizationDetails,
      });
      // Make sure the account is ACTIVE before the first tool call.
      if (connectedAccount?.status !== ConnectorStatus.ACTIVE) {
      ({ connectedAccount } = await scalekit.actions.upsertConnectedAccount({
      connectionName: 'devinmcp',
      identifier: 'user_123',
      authorizationDetails,
      }));
      }
  4. Authorize a user and make your first call

    quickstart.mts
    import { ScalekitClient } from '@scalekit-sdk/node'
    import 'dotenv/config'
    import { createInterface } from 'node:readline/promises'
    const scalekit = new ScalekitClient(
    process.env.SCALEKIT_ENVIRONMENT_URL,
    process.env.SCALEKIT_CLIENT_ID,
    process.env.SCALEKIT_CLIENT_SECRET,
    )
    const actions = scalekit.actions
    const connector = 'devinmcp'
    const identifier = 'user_123'
    // Generate an authorization link for the user
    const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
    console.log('Authorize Devin MCP:', link)
    const rl = createInterface({ input: process.stdin, output: process.stdout })
    await rl.question('Press Enter after authorizing...')
    rl.close()
    // Make your first call
    const result = await actions.executeTool({
    connector,
    identifier,
    toolName: 'devinmcp_list_wiki_repos',
    toolInput: {},
    })
    console.log(result)
    Terminal window
    npx tsx quickstart.mts

    Each user opens the link once and enters their Devin MCP server credentials there. If your app already has a user's credentials, add the account from your backend instead, as the console steps above show. See Authorize a user for the full flow and statuses.

Tools

Pass the exact name to execute_tool
Try in PlaygroundRequest a tool
  • devinmcp_list_wiki_reposList the repositories that have a DeepWiki index, i.e.Read-only

    List Wiki Repos

    List the repositories that have a DeepWiki index, i.e. the ones that ask_wiki_question and read_wiki_* can answer about. This is not the set of repositories you have access to: repositories without a generated wiki are omitted.

    Inputs

    This tool takes no inputs.

  • devinmcp_ask_wiki_questionAsk any question about a GitHub repository's codebase and get an AI-powered answer grounded in its DeepWiki.Write

    Ask Wiki Question

    Ask any question about a GitHub repository's codebase and get an AI-powered answer grounded in its DeepWiki.

    Inputs

    questionstringrequired
    The question to ask about the repository.
    repoNamestringrequired
    GitHub repository or list of repositories (max 10) in owner/repo format.
  • devinmcp_devin_automation_manageManage Devin automations that run Devin in response to events (GitHub, Slack, Linear, schedules, webhooks) — list, get, create, update, delete, or run them.Write

    Devin Automation Manage

    Manage Devin automations that run Devin in response to events (GitHub, Slack, Linear, schedules, webhooks) — list, get, create, update, delete, or run them.

    Inputs

    actionstringrequired
    Call 'schemas' before building or changing an automation. 'validate_*' dry-run without approval. 'run' fires now and bypasses trigger conditions. 'script_check_*' work only when 'schemas' reports Preflight available.one of listgetschemasmanaged_webhookstemplatescreatevalidate_createupdatevalidate_updateset_avatardeleterunrun_grantscript_check_getscript_check_samplesscript_check_eventsscript_check_eventscript_check_testscript_check_run
    actionsarray
    Required for 'create'; replaces all actions on 'update'. Shape in action='schemas'; workflow in the managing-automations skill, if available.
    afterstring
    Cursor from a previous page's end_cursor.
    agent_typesstring
    Filter 'list': new_session, long_running, triage_session, auto_triage, code_scan, remediate_finding, incident_session.
    attachment_urlstring
    For set_avatar: the image URL returned by upload_attachment. Sets only your own bot's public avatar.
    automation_idstring
    Target automation (`auto-<uuid>` or a bare UUID). Omit for 'list', 'schemas', 'managed_webhooks', 'templates', 'create', and draft script checks.
    backgroundstring
    For set_avatar: background color (black, brown, red, orange, yellow, green, teal, blue, purple, pink, gray, rose, magenta, indigo, azure, emerald, tan, amber, tangerine, vermilion, silver, charcoal) or transparent. Omit attachment_url to change only the current custom image's background.
    concurrencyobject
    No description.
    creator_idstring
    Filter 'list' by creator user id.
    descriptionstring
    'create' only: one sentence of at most 70 characters, starting with a verb, saying what it does and where the result goes, not when it runs. Shown on the approval card; not saved.
    enabledstring
    Sets the state on 'create'/'update'; filters on 'list'.
    event_datastring
    'run' only: structured stand-in for the event.
    event_typestring
    Filter 'list' by trigger event type; required for 'run_grant'.
    eventsstring
    'run' only: up to 50 source records, one run each.
    firststring
    Page size for 'list' (default 100) and 'script_check_events' (default 20); max 200.
    grant_idstring
    'run' with events: from 'run_grant'.
    limitsobject
    No description.
    llm_checkobject
    LLM check.
    metadatastring
    String labels; filters on 'list'.
    namestring
    1-500 characters. Required for 'create' and 'validate_create'.
    notificationsobject
    No description.
    promptstring
    'run' only: free-text stand-in for the event (max 10000 chars).
    queued_event_idstring
    For 'script_check_event' and 'script_check_test'; from 'script_check_events'.
    reasonstring
    'update' and 'delete' only: one sentence of at most 100 characters saying why. Shown on the approval card; not saved.
    report_required_permissionsboolean
    Set by the pre-approval hook; not needed by hand.default false
    run_asobject
    Required for 'create'.
    run_idstring
    For 'script_check_run': from 'script_check_test'.
    sample_event_typestring
    For 'script_check_samples' and 'script_check_test'.
    sample_webhook_bodystring
    For 'script_check_test' on 'webhook:incoming': the JSON body the webhook will receive.
    script_checkobject
    Preflight check script.
    searchstring
    Filter 'list' by name substring.
    security_profilestring
    No description.
    session_settingsobject
    No description.
    template_idstring
    'create' only: a template_id from action='templates'.
    toolsobject
    Tool grants for spawned sessions.
    triggersarray
    Required for 'create'; replaces all triggers on 'update'. Shape in action='schemas'; workflow in the managing-automations skill, if available.
    ttl_secondsstring
    'run_grant' only: grant lifetime in seconds (default 3600, max 86400).
  • devinmcp_devin_billing_tag_manageManage Devin billing tags (groupings of Devin sessions for usage tracking) with a single action-based tool: list/get/create tags, assign sessions one at a time or in bulk (up to 500 per call, sent in sequential batches of 100), and read the billing tags of up to 100 sessions at once ('get_session_tags').Write

    Devin Billing Tag Manage

    Manage Devin billing tags (groupings of Devin sessions for usage tracking) with a single action-based tool: list/get/create tags, assign sessions one at a time or in bulk (up to 500 per call, sent in sequential batches of 100), and read the billing tags of up to 100 sessions at once ('get_session_tags'). Only available in private mode (via devin.ai endpoints).

    Inputs

    actionstringrequired
    Operation to perform: 'list', 'get', 'create', 'assign_session', 'assign_sessions', or 'get_session_tags'.one of listgetcreateassign_sessionassign_sessionsget_session_tags
    afterstring
    Opaque cursor for 'list' pagination (from a previous response's end_cursor).
    assignmentsstring
    Map of Devin session ID -> billing tag ID (1-500 entries). Required for 'assign_sessions'. A null value marks the session as miscellaneous (no billing tag). Assignments are sent to the API in sequential batches of 100. Partial success: sessions or tags not found in the organization, or sessions you can't write, are reported per entry.
    billing_tag_idstring
    Billing tag id. Required for 'get'. For 'assign_session', the tag to assign; omit it to mark the session as miscellaneous (no tag).
    descriptionstring
    Free-form description of what work the tag covers. Optional for 'create'.
    devin_idstring
    ID of the Devin session to assign. Required for 'assign_session'. Assigning replaces the session's current billing tag.
    firststring
    Page size for 'list' pagination (default 100, max 200).
    namestring
    Billing tag display name. Required for 'create'.
    session_idsstring
    Session ids to look up billing tags for, at most 100. Required for 'get_session_tags'. Both 'devin-abc123' and 'abc123' are accepted. Each session comes back with its billing tag, as having no billing tag when it was categorized as Miscellaneous or its tag was deleted, as never categorized, or as not found.
  • devinmcp_devin_blueprint_testTest a candidate blueprint YAML in an authoring VM: start → run initialize → fix → run initialize again (clean VM) → run maintenance → then persist with update_environment_config.Write

    Devin Blueprint Test

    Test a candidate blueprint YAML in an authoring VM: start → run initialize → fix → run initialize again (clean VM) → run maintenance → then persist with update_environment_config.

    Inputs

    actionstringrequired
    Operation: start, run, get, cancel, or close.one of startrungetcancelclose
    afterinteger
    Log cursor for get.default 0
    author_session_idstring
    Author session ID for run, get, cancel, and close.
    namestring
    Knowledge command name, when target is knowledge.
    platformstring
    Authoring VM platform label.default default
    platformsstring
    Platform labels matching YAML "runs-on" values; use "default" for the default platform. Defaults to [platform].
    poll_interval_secondsinteger
    Seconds between run polls; capped so keepalives keep flowing.default 5
    repo_namestring
    Repository name used to create or set the session working directory.
    run_idstring
    Run ID for get and cancel.
    scopestring
    Optional repository scope for the run working directory.
    targetstring
    Blueprint target: initialize, maintenance, or knowledge.
    wait_timeout_secondsinteger
    Maximum time to wait for a run. Capped below the 15-minute approval window; use `get` to keep following a longer run.default 600
    yamlstring
    Candidate blueprint YAML for a run.
  • devinmcp_devin_campaign_manageRead and publish a bot's campaign: its plan, its append-only readings and its status.Write

    Devin Campaign Manage

    Read and publish a bot's campaign: its plan, its append-only readings and its status. Access follows the bot: a personal bot's owner and the bot's own sessions, or any org member for a multiplayer bot; org admins may read a personal bot's campaign. A Devin session may publish drafts, readings and status, but never approve the plan: send plan.approved exactly as stored (a person approves in the campaign panel). 'running' requires an approved plan. The first update creates the campaign.

    Inputs

    actionstringrequired
    'get' reads the campaign; 'update' publishes the given fields; 'validate' runs every check 'update' makes without saving.one of getupdatevalidate
    automation_idstring
    The bot's root automation (auto-<id>), resolved to its bot. A campaign Devin passes its own session automation id here. Pass this or bot_id.
    bot_idstring
    The bot (bot-<id>) whose campaign to read or publish. Pass this or automation_id.
    campaignstring
    For update/validate: the complete plan and/or complete latest_readings document, plus an optional status. Omitted fields stay unchanged. Parse local YAML into JSON first.
  • devinmcp_devin_code_scan_manageManage Devin code scans and scan profiles — list scans and findings, manage profiles, create scans, or remediate findings.Write

    Devin Code Scan Manage

    Manage Devin code scans and scan profiles — list scans and findings, manage profiles, create scans, or remediate findings.

    Inputs

    actionstringrequired
    The operation to perform. Per-action rules are in the tool description and the creating-code-scans skill. A 409 means the scan's state does not allow it.one of list_scansget_scanlist_findingslist_threatslist_scan_typeslist_profilesget_profilecreate_profilecreate_profile_setupdate_profilecreate_scancreate_bulk_scanget_bulk_scanupdate_scanscan_new_commitsrecover_scancancel_scancreate_auto_scandelete_auto_scanupdate_finding_statusupdate_finding_contentget_sessionremediate_findingremediate_findingsunlink_finding_session
    afterstring
    Opaque cursor for list pagination (from a previous response's end_cursor). 'get_bulk_scan': the repository name the previous page says to continue after.
    attachment_urlsstring
    Devin attachment URLs for 'create_scan' (at most 10), e.g. findings for an ingest scan.
    batch_idstring
    Required for 'get_bulk_scan': the batch ID returned by 'create_bulk_scan'.
    board_idstring
    'create_scan': finding board named by the setup. Never invent one.
    branchstring
    Base branch for remediation pull requests; omit to use the repository default branch.
    commit_shastring
    Commit to scan first ('create_scan', single repository); omit for the default branch.
    communication_guidancestring
    What happens when the scan finishes ('create_profile', 'update_profile'); discover scans only.
    criteriastring
    'create_profile_set' only: every criterion parsed from the catalog, each once ({'id', 'text', 'title', 'guidance'}).
    descriptionstring
    Description of the profile ('create_profile' and 'update_profile'), or the finding's new description ('update_finding_content').
    dismissal_notestring
    Dismissal note for 'update_finding_status'; stored only when dismissed.
    dismissal_reasonstring
    'update_finding_status' when dismissed: 'false_positive', 'mitigated', 'accepted_risk', 'duplicate', 'not_actionable'.
    dry_runboolean
    'create_bulk_scan': list the repositories it would scan without starting anything.default false
    effortstring
    'normal' or 'deep' ('create_scan', 'create_bulk_scan', 'create_auto_scan'). Only security scans offer the choice. Describe it by time and thoroughness, never cost.
    enabledboolean
    Whether the 'create_auto_scan' schedule starts enabled (default true).default true
    exclude_globsstring
    Glob patterns of files to exclude from scans ('create_profile', 'update_profile'; 'create_scan' without a profile).
    filter_namestring
    'create_bulk_scan' with dry_run=true: list the repositories whose name contains this text. Start the scans with the approved names in repos.
    finding_idstring
    The finding ID for 'remediate_finding', 'update_finding_status' or 'unlink_finding_session'.
    finding_idsstring
    Findings of scan_id for 'remediate_findings', 'update_finding_status' or 'unlink_finding_session' (1-1000).
    firststring
    Page size for list actions (default 100, max 200).
    fix_effortstring
    The finding's new fix effort ('update_finding_content').
    hoststring
    Git host of the repository/repositories ('create_scan' only).
    impactstring
    The finding's new impact ('update_finding_content').
    include_globsstring
    Glob patterns of files to include in scans ('create_profile', 'update_profile'; 'create_scan' without a profile).
    ingestion_source_guidancestring
    Guidance describing the ingestion source ('create_profile' and 'update_profile').
    interactivestring
    'create_scan', security scans only: true pauses for the user to review the threat model; false (default) runs unattended.
    investigation_guidancestring
    Guidance for the investigation phase ('create_profile' and 'update_profile').
    localestring
    BCP-47 tag of the user's language ('create_scan', 'create_bulk_scan'). Pass it when they are not writing in English.
    modestring
    Profile mode: 'discover' (default) or 'ingest' ('create_profile' only; the mode is fixed at creation).
    notestring
    The finding's new note; empty clears it ('update_finding_content').
    ownersstring
    The finding's code owners; empty clears them ('update_finding_content').
    post_ingestion_guidancestring
    Guidance for the post-ingestion phase ('create_profile' and 'update_profile').
    profile_groupsstring
    'create_profile_set' only: one entry per profile, together covering every criterion in 'criteria' exactly once, at most 50 each.
    profile_idstring
    Required for 'get_profile', 'update_profile', 'update_scan'; optional for 'create_scan', 'create_bulk_scan'. Required for non-security scans.
    profile_namestring
    Name of the profile. Required for 'create_profile'; optional new name for 'update_profile'.
    promptstring
    'create_scan': text added to the scan's kickoff message, e.g. an '@skills:' mention from the user's request and what to use it for. Only a space or line break may follow a skill's name: a period there is read as part of it.
    querystring
    Text search: over finding title, description, and file path ('list_findings', combines with repo_name and other filters), or over a threat's id, title, description and tags, such as a CVE id or package name ('list_threats').
    recommendationstring
    The finding's new recommendation; empty clears it ('update_finding_content').
    related_finding_idsstring
    The finding's related finding IDs; empty clears them ('update_finding_content').
    remediation_guidancestring
    Guidance for the remediation phase ('create_profile' and 'update_profile').
    repo_namestring
    Repository filter, or the repository to scan ('create_scan'; exactly one of repo_name or repos).
    report_guidancestring
    Guidance for the report phase ('create_profile' and 'update_profile').
    reposstring
    'create_scan': repositories to scan together (exactly one of repo_name or repos). 'create_bulk_scan': repositories to scan separately, one scan each (at most 1000; repos or filter_name).
    resolution_notestring
    Resolution note for 'update_finding_status'; stored only when resolved.
    rrulestring
    Required for 'create_auto_scan': iCal RRULE driving the recurring schedule for scanning new commits, e.g. 'FREQ=DAILY'.
    scan_idstring
    The scan the action applies to.
    scan_pathsstring
    Repository-relative paths to restrict scans to ('create_profile', 'update_profile'; 'create_scan' without a profile); [] lifts it.
    scan_typestring
    Scan type from 'list_scan_types' ('create_scan', 'create_bulk_scan', 'create_profile', 'update_profile', 'list_profiles', 'list_scans'). Must match the profile's type.
    select_allboolean
    'remediate_findings' instead of finding_ids: every open finding of scan_id, narrowed by severity and repo_name. Fails above 1000.default false
    severitystring
    Filter findings to these severities: 'critical', 'high', 'medium', 'low' ('list_findings', and 'remediate_findings' with select_all), or the finding's new severity ('update_finding_content').
    statusstring
    Filter for 'list_findings', or the new status for 'update_finding_status': 'open', 'dismissed', 'resolved'.
    strategystring
    Scan strategy ('create_scan' only): 'agentic_map_reduce' (default) or 'exploit' (security scans only). Non-default strategies are staging/dev-only.
    summarystring
    One-sentence, user-facing description of the proposed change, shown on the approval card when 'update_profile' targets a 'security' profile.
    threat_model_guidancestring
    Guidance describing the threat model ('create_profile' and 'update_profile').
    time_afterstring
    Unix seconds lower bound. Optional creation-time filter ('list_scans' only).
    time_beforestring
    Unix seconds upper bound. Optional creation-time filter ('list_scans' only).
    titlestring
    The finding's new title ('update_finding_content').
    triage_guidancestring
    Guidance for the triage phase ('create_profile' and 'update_profile').
    type_configstring
    Compliance profiles only: {'system', 'framework', 'criteria'}. Copy 'system.guidance' unchanged. For 'create_profile_set', only the shared 'system' and 'framework'. See the creating-code-scans skill.
    validation_guidancestring
    Guidance for the validation phase ('create_profile' and 'update_profile').
    validation_severitiesstring
    Severities runtime validation covers ('create_profile', 'update_profile'; security profiles only).
  • devinmcp_devin_find_settingFind a Devin webapp setting and get a deep-link URL to it.Write

    Devin Find Setting

    Find a Devin webapp setting and get a deep-link URL to it. Use for "where do I change X?". The output includes everything needed to build the final URL.

    Inputs

    querystring
    Feature keywords (e.g. "queue messages", "fast mode"). Omit for the full settings catalog.