# TraceCtrl

> Security Observability & Control for Agentic AI

## Docs

- [For AI Agents](https://docs.tracectrl.ai/agents.md): Copy-paste instructions to give any AI assistant full context on TraceCtrl.
- [Guardrails For Agent](https://docs.tracectrl.ai/api-reference/agent-guardrails.md): Guardrails registered against a specific agent
- [Agent Risk Scores](https://docs.tracectrl.ai/api-reference/agent-scores.md): Per-agent risk scores
- [Get Agent Tools](https://docs.tracectrl.ai/api-reference/agent-tools.md): Tools used by a specific agent
- [List Agents (Inventory)](https://docs.tracectrl.ai/api-reference/agents.md): Full agent inventory with tool counts and prompt metadata
- [Attack Graph Overlay](https://docs.tracectrl.ai/api-reference/attack-graph-overlay.md): Compromised nodes and attack edges overlayed on the topology
- [Attack Graph Paths](https://docs.tracectrl.ai/api-reference/attack-graph-paths.md): Detailed attack paths for the findings panel
- [Attack Paths](https://docs.tracectrl.ai/api-reference/attack-paths.md): Ranked attack paths from TAGAAI rules
- [Guardrail Invocations](https://docs.tracectrl.ai/api-reference/guardrail-invocations.md): Recent guardrail evaluation spans
- [List Guardrails](https://docs.tracectrl.ai/api-reference/guardrails-list.md): Registered guardrails across the system
- [Health Check](https://docs.tracectrl.ai/api-reference/health.md): Check if the engine is running
- [API Overview](https://docs.tracectrl.ai/api-reference/overview.md): TraceCtrl Engine REST API reference
- [List Projects](https://docs.tracectrl.ai/api-reference/projects.md): Returns distinct service names seen in trace data
- [Protector Plus Config](https://docs.tracectrl.ai/api-reference/protector-config.md): Read, update, and test the TraceCtrl Guards (Protector Plus) configuration
- [List Agents](https://docs.tracectrl.ai/api-reference/risk-agents.md): Returns all discovered agents with inventory data
- [Risk Summary](https://docs.tracectrl.ai/api-reference/risk-summary.md): System-wide risk summary
- [Apply Scan Fixes](https://docs.tracectrl.ai/api-reference/scan-fix.md): Apply automated remediations to openclaw.json
- [Scan Job Status](https://docs.tracectrl.ai/api-reference/scan-status.md): Poll for the status of a triggered scan
- [Trigger Scan](https://docs.tracectrl.ai/api-reference/scan-trigger.md): Start a scan job and poll for completion
- [Validate Workspace Path](https://docs.tracectrl.ai/api-reference/scan-validate-path.md): Check whether a filesystem path is a valid OpenClaw workspace
- [Get Scan](https://docs.tracectrl.ai/api-reference/scans-detail.md): Results and topology for a specific scan
- [Latest Scan](https://docs.tracectrl.ai/api-reference/scans-latest.md): Most recent scan results with config drift detection
- [List Scans](https://docs.tracectrl.ai/api-reference/scans-list.md): All stored scans with summary counts
- [Upload Scan Results](https://docs.tracectrl.ai/api-reference/scans-upload.md): Store scan results produced by a remote TraceCtrl CLI
- [List Sessions](https://docs.tracectrl.ai/api-reference/sessions.md): Returns recent traces grouped as sessions
- [Latest Trace Spans](https://docs.tracectrl.ai/api-reference/sessions-latest-spans.md): Spans from the most recent trace
- [Get Trace Spans](https://docs.tracectrl.ai/api-reference/sessions-spans.md): Returns all spans for a specific trace
- [Get Agent Detail](https://docs.tracectrl.ai/api-reference/topology-agent.md): Returns details for a specific agent
- [Get Topology Graph](https://docs.tracectrl.ai/api-reference/topology-graph.md): Returns the full agent-tool topology graph
- [List Violations](https://docs.tracectrl.ai/api-reference/violations.md): Guardrail violations with optional filtering
- [Recent Violations](https://docs.tracectrl.ai/api-reference/violations-recent.md): Convenience endpoint for the most recent violations
- [Stream Violations (SSE)](https://docs.tracectrl.ai/api-reference/violations-stream.md): Server-Sent Events stream of guardrail violations
- [Architecture](https://docs.tracectrl.ai/architecture.md): How TraceCtrl's components fit together
- [TraceCtrl Bootcamp (Openclaw)](https://docs.tracectrl.ai/bootcamp/guide.md): A hands-on workshop — from zero to full security observability for AI agents in 2 hours
- [Prerequisites](https://docs.tracectrl.ai/bootcamp/prerequisites.md): System requirements, software, and API keys to set up before the TraceCtrl bootcamp. Complete this before the session.
- [TraceCtrl Bootcamp (Strands)](https://docs.tracectrl.ai/bootcamp/strands-guide.md): Hands-on workshop — instrument an AWS Strands agent with TraceCtrl and go from zero traces to full security observability in 90 minutes.
- [Agno](https://docs.tracectrl.ai/instrumentors/agno.md): Instrument Agno agents with TraceCtrl
- [CrewAI](https://docs.tracectrl.ai/instrumentors/crewai.md): Instrument CrewAI agents and crews with TraceCtrl
- [Google ADK](https://docs.tracectrl.ai/instrumentors/google-adk.md): Instrument Google Agent Development Kit agents with TraceCtrl
- [LangChain / LangGraph](https://docs.tracectrl.ai/instrumentors/langchain.md): Instrument LangChain and LangGraph agents with TraceCtrl
- [OpenClaw](https://docs.tracectrl.ai/instrumentors/openclaw.md): Trace OpenClaw gateway sessions with TraceCtrl
- [AWS Strands](https://docs.tracectrl.ai/instrumentors/strands.md): Instrument AWS Strands agents with TraceCtrl
- [Introduction](https://docs.tracectrl.ai/introduction.md): Security observability and control for agentic AI systems
- [Dashboard](https://docs.tracectrl.ai/platform/dashboard.md): React dashboard for visualizing agents, traces, alerts, guardrails, and attack paths
- [Data Flow](https://docs.tracectrl.ai/platform/data-flow.md): How spans flow from your agent to the dashboard
- [Intelligence Engine](https://docs.tracectrl.ai/platform/engine.md): FastAPI backend that processes spans into agent intelligence
- [Quickstart](https://docs.tracectrl.ai/quickstart.md): Get TraceCtrl running end-to-end in 5 minutes
- [Configuration](https://docs.tracectrl.ai/sdk/configuration.md): Configure the TraceCtrl SDK via code or environment variables
- [Context Propagation](https://docs.tracectrl.ai/sdk/context-propagation.md): Correlate traces across multiple services and agents
- [Guardrails](https://docs.tracectrl.ai/sdk/guardrails.md): Declare LLM-judge guardrails on agent inputs and outputs
- [Installation](https://docs.tracectrl.ai/sdk/installation.md): Install the TraceCtrl SDK and framework instrumentors
- [Protector Plus](https://docs.tracectrl.ai/sdk/protector-plus.md): External LLM firewall integration via tracectrl.guard()
- [Session Management](https://docs.tracectrl.ai/sdk/session-management.md): Correlate spans across a user session or conversation
- [Security Enrichment](https://docs.tracectrl.ai/security/enrichment.md): How TraceCtrlSpanProcessor enriches spans with security context
- [Span Schema](https://docs.tracectrl.ai/security/span-schema.md): Complete reference for TraceCtrl span attributes
- [Tool Categories](https://docs.tracectrl.ai/security/tool-categories.md): Automatic risk classification for every tool call

## OpenAPI Specs

- [openapi](https://docs.tracectrl.ai/api-reference/openapi.json)