Waxell

aid: waxell
url: https://raw.githubusercontent.com/api-evangelist/waxell/refs/heads/main/apis.yml
name: Waxell
type: Index
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
description: >-
  Waxell is an AI agent governance and observability platform that provides
  runtime policy enforcement, auto-instrumented LLM telemetry, MCP governance,
  cost management, and durable workflow execution for agents built in any
  Python framework or third-party agentic tool (Claude Code, Cursor,
  LangChain, CrewAI, OpenAI Agents SDK, and 200+ more).
created: '2026-05-06'
modified: '2026-05-06'
specificationVersion: '0.19'
tags:
  - AI Agent Governance
  - Observability
  - Policy Enforcement
  - LLM Telemetry
  - Cost Management
  - MCP
  - Agent Runtime
url-source: https://waxell.ai/
apis:
  - aid: waxell:observe
    name: Waxell Observe API
    tags:
      - AI Agent Governance
      - Observability
      - LLM Telemetry
      - Policy Enforcement
      - Cost Management
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    baseURL: https://{tenant}.waxell.dev/api/v1/observe
    humanURL: https://waxell.ai/docs/observe/api/endpoints
    description: >-
      The Waxell Observe REST API exposes the AI agent governance and
      observability control plane. It is used by the waxell-observe Python
      SDK and the Developer MCP server to record runs, log LLM calls, spans,
      steps and scores, evaluate runtime governance policies, manage prompts,
      and administer the model cost catalog. Endpoints live under
      /api/v1/observe/ on a tenant-specific *.waxell.dev host and accept the
      same wax_sk_ keys via X-Wax-Key or Authorization: Bearer.
    properties:
      - type: Documentation
        url: https://waxell.ai/docs
      - type: APIReference
        url: https://waxell.ai/docs/observe/api/endpoints
      - type: Quickstart
        url: https://waxell.ai/docs/observe/quickstart
      - type: Authentication
        url: https://waxell.ai/docs/observe/api/authentication
      - type: SDK
        url: https://pypi.org/project/waxell-observe/
      - type: OpenAPI
        url: openapi/waxell-observe-openapi.yml
      - type: JSONSchema
        url: json-schema/waxell-run-schema.json
      - type: JSONSchema
        url: json-schema/waxell-llm-call-schema.json
      - type: JSONSchema
        url: json-schema/waxell-policy-decision-schema.json
      - type: JSONSchema
        url: json-schema/waxell-span-schema.json
      - type: JSONStructure
        url: json-structure/waxell-run-structure.json
      - type: JSONStructure
        url: json-structure/waxell-policy-decision-structure.json
  - aid: waxell:developer-mcp
    name: Waxell Developer MCP Server
    tags:
      - MCP
      - AI Agent Governance
      - Developer Tools
      - Coding Agents
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    baseURL: https://dev-mcp.waxell.dev
    humanURL: https://waxell.ai/docs/agents/overview
    description: >-
      Waxell Developer MCP is a hosted Model Context Protocol server that
      lets coding agents (Claude Code, Cursor, Windsurf, VS Code, Claude
      Desktop) query a Waxell instance in real time. It exposes 15 live
      tools (agent health, error investigation, LLM cost tracking,
      governance policy review, account signup) plus 8 documentation
      resources at waxell://docs/*. Connection is SSE; per-client
      authentication uses Bearer tokens in the Authorization header.
    properties:
      - type: Documentation
        url: https://waxell.ai/docs/agents/overview
      - type: APIReference
        url: https://dev-mcp.waxell.dev/sse
      - type: Authentication
        url: https://waxell.ai/docs/agents/overview
      - type: GitHubRepository
        url: https://gitlab.com/waxell/agentforge
common:
  - type: Documentation
    url: https://waxell.ai/docs
  - type: GettingStarted
    url: https://waxell.ai/docs/observe/quickstart
  - type: Quickstart
    url: https://waxell.ai/docs/observe/quickstart
  - type: SDK
    url: https://waxell.ai/docs/observe/quickstart
  - type: Console
    url: https://waxell.dev
  - type: SignUp
    url: https://waxell.ai/get-access
  - type: Pricing
    url: https://waxell.ai/get-access
  - type: Plans
    url: plans/waxell-plans-pricing.yml
  - type: RateLimits
    url: rate-limits/waxell-rate-limits.yml
  - type: FinOps
    url: finops/waxell-finops.yml
  - type: StatusPage
    url: https://status.waxell.dev
  - type: Blog
    url: https://waxell.ai/blog
  - type: Glossary
    url: https://waxell.ai/glossary
  - type: Security
    url: https://waxell.ai/docs/security
  - type: TrustCenter
    url: https://app.vanta.com/callsine.com/trust/pg7qc55eh5ge6ejjv7zxksy
  - type: Compliance
    url: https://waxell.ai/docs/security
  - type: LinkedIn
    url: https://www.linkedin.com/company/waxell-ai
  - type: GitHubRepository
    url: https://gitlab.com/waxell/agentforge
  - type: SpectralRules
    url: rules/waxell-rules.yml
  - type: Vocabulary
    url: vocabulary/waxell-vocabulary.yml
  - type: NaftikoCapability
    url: capabilities/agent-governance.yaml
  - type: NaftikoCapability
    url: capabilities/shared/observe.yaml
  - type: JSON-LD
    url: json-ld/waxell-context.jsonld
  - type: Example
    url: examples/waxell-start-run-example.json
  - type: Example
    url: examples/waxell-record-llm-call-example.json
  - type: Example
    url: examples/waxell-policy-check-example.json
  - type: Example
    url: examples/waxell-get-prompt-example.json
  - type: Features
    data:
      - name: Auto-Instrumentation
        description: Two-line setup auto-instruments 200+ libraries (OpenAI, Anthropic, LangChain, LlamaIndex, CrewAI, LiteLLM, etc.) without code changes.
      - name: Runtime Policy Enforcement
        description: 26 policy categories (cost, kill switch, PII, compliance, scope, safety) returning seven decisions (allow, warn, redact, throttle, block, skip, retry).
      - name: MCP Governance
        description: Auto-instrumentor, server middleware, and governance proxy for Model Context Protocol traffic with PII scanning and rug-pull detection.
      - name: Cost Management
        description: Built-in model pricing for 20+ models, tenant overrides via REST, budget enforcement that warns/throttles/blocks at thresholds.
      - name: Prompt Management
        description: Versioned managed prompts retrievable by name and label (e.g. production, staging) directly from the SDK.
      - name: Workflow Envelope
        description: Durable execution boundary with checkpoint and resume; Redis-backed in production, in-memory for development.
      - name: Human-in-the-Loop Approvals
        description: Custom handlers route policy blocks to Slack, webhooks, or terminal prompts for human review.
      - name: Audit Trail
        description: Immutable, timestamped record of all agent actions, decisions, and governance events.
      - name: Developer MCP
        description: Hosted SSE MCP server (dev-mcp.waxell.dev/sse) with 15 live tools and 8 docs resources for coding agents.
      - name: Field-Level Encryption
        description: PII fields encrypted at the application layer with AES-256-GCM and AWS KMS (FIPS 140-2 Level 3) before database storage.
    sources:
      - https://waxell.ai/
      - https://waxell.ai/docs
      - https://waxell.ai/docs/security
      - https://waxell.ai/compare
    updated: '2026-05-06'
  - type: UseCases
    data:
      - name: Govern Third-Party Coding Agents
        description: Enforce policies on Claude Code, Cursor, Windsurf, VS Code, and Claude Desktop without modifying their code.
      - name: Instrument Self-Built Agents
        description: Add full observability to LangChain, CrewAI, OpenAI Agents SDK, or custom Python agents with the @waxell.observe decorator.
      - name: Cost-Capped Agent Deployment
        description: Set budgets on token spend per agent, user, or tenant; block runs that exceed configured limits.
      - name: PII-Safe MCP Tool Use
        description: Scan MCP tool inputs/outputs for PII, credentials, and secrets with warn/block/redact responses.
      - name: Durable Long-Running Workflows
        description: Use the WorkflowEnvelope to checkpoint multi-step agent workflows so they can resume after interruption.
      - name: Compliance-Ready AI Operations
        description: Maintain SOC 2 Ready posture with immutable audit trails, encrypted PII, and EU data residency.
    sources:
      - https://waxell.ai/
      - https://waxell.ai/docs
    updated: '2026-05-06'
  - type: Integrations
    data:
      - name: OpenAI
        description: Auto-instrumented LLM provider; cost and token tracking out of the box.
      - name: Anthropic
        description: Auto-instrumented LLM provider; supports Claude family models.
      - name: LangChain / LangGraph
        description: First-class callback handler (WaxellLangChainHandler) for tracing chains and graphs.
      - name: CrewAI
        description: Auto-instrumented multi-agent framework support.
      - name: LlamaIndex
        description: Tracing for RAG pipelines built with LlamaIndex.
      - name: LiteLLM
        description: Unified telemetry across LiteLLM-routed providers.
      - name: Claude Code
        description: Governance overlay for Anthropic's Claude Code coding agent via the Developer MCP.
      - name: Cursor / Windsurf / VS Code
        description: Coding-agent governance via the SSE MCP server at dev-mcp.waxell.dev.
      - name: OpenAI Agents SDK
        description: Auto-instrumentation for OpenAI's Agents SDK runs.
      - name: AWS Bedrock / Azure OpenAI / Google Vertex AI
        description: Cloud LLM providers covered by Waxell's auto-instrumentation.
      - name: Pinecone / Weaviate / Qdrant / Milvus / Chroma
        description: Vector database call tracing with retrieval span recording.
      - name: Slack / Webhooks
        description: Human-in-the-loop approval handlers for policy blocks.
      - name: Stripe
        description: Listed subprocessor for billing.
    sources:
      - https://waxell.ai/docs/integrations
    updated: '2026-05-06'
  - type: Solutions
    data:
      - name: Connect
        description: Govern third-party agents (Claude Code, Cursor) without code changes via the MCP governance proxy.
      - name: Observe
        description: Instrument self-built agents with auto-instrumentation, policy enforcement, and cost attribution.
      - name: Runtime
        description: Governed execution environment for high-risk workflows with the durable WorkflowEnvelope.
    sources:
      - https://waxell.ai/
    updated: '2026-05-06'
maintainers:
  - FN: Kin Lane
    email: [email protected]
    url: https://apievangelist.com