Gazebo
    ServicesAgentsDocsWritingPricing
    Log inSign up
    Log in

    Getting Started

    • Introduction
    • How Gazebo works
    • Core concepts
    • Quickstart

    Agent Access

    • Access profiles
    • Bearer tokens
    • Audit logs

    MCP

    • Overview
    • Cursor
    • Claude Desktop
    • Claude Code
    • OpenAI / Codex
    • Devin / Windsurf
    • Any MCP client
    • Get credential reference

    Authorization

    • Agent Consent Flow

    REST API

    • Overview
    • Authentication
    • Services
    • Agents
    • Access policies
    • Audit events
    • Consent administration
    • Agent credential access
    • Errors and limits

    Developer Tools

    • TypeScript SDK
    • CLI

    Security

    • Overview
    • Encryption model
    • Credential storage
    • Audit and compliance
    • Responsible disclosure
    Documentation

    Getting Started

    • Introduction
    • How Gazebo works
    • Core concepts
    • Quickstart

    Agent Access

    • Access profiles
    • Bearer tokens
    • Audit logs

    MCP

    • Overview
    • Cursor
    • Claude Desktop
    • Claude Code
    • OpenAI / Codex
    • Devin / Windsurf
    • Any MCP client
    • Get credential reference

    Authorization

    • Agent Consent Flow

    REST API

    • Overview
    • Authentication
    • Services
    • Agents
    • Access policies
    • Audit events
    • Consent administration
    • Agent credential access
    • Errors and limits

    Developer Tools

    • TypeScript SDK
    • CLI

    Security

    • Overview
    • Encryption model
    • Credential storage
    • Audit and compliance
    • Responsible disclosure
    DocsSecurityAudit and compliance

    Audit and compliance

    Every credential access attempt is logged — whether it succeeds, is denied, or fails due to a missing credential. Logs are append-only.

    These are credential-retrieval records, not a record of subsequent provider actions. The method is the agent's declared intent. Token authentication failures do not create credential-access events; use provider logs for downstream investigation.

    What gets logged

    EventLoggedNotes
    get_credential — successYesCredential name returned to agent
    get_credential — denied (method)YesMethod not in access profile
    get_credential — denied (service)YesService not in access profile
    get_credential — not foundYesNo credential stored for service
    get_identityNoRead-only; no credential access
    list_audit_eventsNoReading logs doesn't create a log entry
    Bearer token validation failureNoOnly credential access is logged

    Log fields

    {
      "id": "evt_...",
      "agentId": "...",
      "agentName": "cursor-dev",
      "service": "stripe",
      "keyName": "primary",
      "action": "read",
      "method": "POST",
      "outcome": "success",
      "createdAt": "2026-07-27T10:23:41.000Z"
    }
    

    The example uses REST account audit-event fields. method is the method declared during retrieval, not an observed provider operation. Outcomes include success, denied and not_found; do not assume a separate denial-reason field is available.

    Append-only

    There is no individual audit-event deletion action in the dashboard or supported public API. This is not a claim of tamper-proof database storage, permanent retention or certification. Apply your account's retention policy and keep independent exports where required.

    Retention

    Do not assume lifetime retention. Check the retention policy for your account and plan, and export records needed for your own retention requirements.

    Accessing logs

    • Dashboard — the Logs page with filtering by agent, service, outcome, and date range
    • MCP — list_audit_events({ limit, service }) returns recent events for the calling agent
    • REST API — GET /api/v1/audit-events with an account token (paginated account credential-access events)

    Compliance use cases

    For compliance workflows that require log export, contact support. Enterprise plans include structured log export via API.

    Explore Gazebo

    Gazebo gives each agent an independent access boundary around your agent credentials, with controls for approval, auditing, and revocation.

    Scoped identities for AI agentsGive every agent only the credentials it needs.Credential security and audit controlsSee how access is protected, recorded, and revoked.
    Credential storageResponsible disclosure
    Gazebo

    Credential management for AI agents. Scoped access, approval controls, and complete audit logs—without hardcoding or rotating shared API keys.

    Gazebo

    • Platform
    • Demo
    • MCP
    • Pricing

    Explore

    • Services
    • Agents
    • Vaults

    Content

    • Writing
    • Topics
    • Blog
    • Docs

    Company

    • About
    • [email protected]

    © 2026 Gazebo. All rights reserved.

    PrivacyTermsSecurity