AI Agent Guardrails (HITL Approval)
# 🛡️ AI Agent Guardrails (HITL Approval) Welcome to the ultimate enterprise safety middleware for autonomous AI agents. When building AI agents using frameworks like **LangChain**, **AutoGPT**, or **CrewAI**, agents often require access to execute sensitive actions—such as deleting files, modifying production databases, or sending customer emails. **AI Agent Guardrails** provides a…
AI Agent Guardrails (HITL Approval) endpoints
| Method | Endpoint | Description |
|---|---|---|
| v1 | ||
| GET |
/v1/session/{sessionId} /v1/session/{sessionId} |
Retrieve the current status and details of a validation session. |
| POST |
/v1/confirm/{sessionId} /v1/confirm/{sessionId} |
Submit a final decision (ACK) for a validation session. |
| POST |
/v1/validate /v1/validate |
Accepts a list of proposed actions, performs Wasm-based preprocessing, and returns a session with 5 random samples for verification. |
AI Agent Guardrails (HITL Approval) pricing
| Plan | Price | Rate limit | Quotas |
|---|---|---|---|
| BASIC | Free | — |
|
| PRO | $9 / month | — |
|
| ULTRA | $29 / month | — |
|
| MEGA | $99 / month | — |
|