Interactive Validators

Use these tools to validate your implementations against the Agent Control Layer specifications.

Validate an Agent Run against the ADP-1 Schema.

{
  "version": "adp-1",
  "run_id": "e7e5c9a4-1c6f-4e4e-9a9a-52f9f0d7e0f1",
  "tenant_id": "tenant-123",
  "trace": {
    "trace_id": "0af7651916cd43dd8448eb211c80319c",
    "span_id": "b7ad6b7169203331"
  },
  "agent": {
    "agent_id": "security-reviewer",
    "agent_version": "v2.1",
    "framework": "acl",
    "aip": {
      "cert_fingerprint": "sha256:a3b9c8d7e6f5a4b3c2d1e0f9a8b7c6d5e4f3a2b1c0d9e8f7a6b5c4d3e2f1a0b9",
      "tenant_id": "tenant-123",
      "capabilities": [
        "agent:security-reviewer",
        "tenant:tenant-123",
        "perm:workflows:read",
        "perm:files:read"
      ]
    }
  },
  "context": {
    "task_description": "Review the submitted code for security vulnerabilities",
    "workflow_key": "security_review_workflow",
    "user_id": "user-456",
    "session_id": "sess-789",
    "labels": [
      "security",
      "code-review",
      "automated"
    ]
  },
  "steps": [
    {
      "index": 0,
      "timestamp": "2025-12-14T10:00:01.234Z",
      "action": {
        "type": "tool_call",
        "name": "read_file",
        "input": {
          "path": "/src/auth/login.ts"
        }
      },
      "observation": {
        "type": "tool_result",
        "output": {
          "content": "// Login handler code...",
          "size_bytes": 2048
        }
      },
      "reflection": {
        "thought": "I have retrieved the login handler. I should analyze it for common vulnerabilities like SQL injection and improper input validation.",
        "next_action_hint": "Analyze code for security issues",
        "uncertainty": 0.15
      },
      "metadata": {
        "model": "gpt-4.1",
        "prompt_tokens": 150,
        "completion_tokens": 45,
        "latency_ms": 320
      }
    },
    {
      "index": 1,
      "timestamp": "2025-12-14T10:00:05.678Z",
      "action": {
        "type": "model_inference",
        "name": "security_analysis",
        "input": {
          "code": "// Login handler code...",
          "analysis_type": "vulnerability_scan"
        }
      },
      "observation": {
        "type": "tool_result",
        "output": {
          "vulnerabilities_found": 1,
          "issues": [
            {
              "severity": "medium",
              "type": "input_validation",
              "line": 42,
              "description": "User input not sanitized before database query"
            }
          ]
        }
      },
      "reflection": {
        "thought": "Found a potential input validation issue. I should include this in my report and suggest a fix.",
        "next_action_hint": "Generate security report",
        "uncertainty": 0.1
      },
      "metadata": {
        "model": "gpt-4.1",
        "prompt_tokens": 890,
        "completion_tokens": 234,
        "latency_ms": 1250
      }
    },
    {
      "index": 2,
      "timestamp": "2025-12-14T10:00:08.901Z",
      "action": {
        "type": "tool_call",
        "name": "policy_judge",
        "input": {
          "draft_output": "Security Review Report: Found 1 medium-severity vulnerability...",
          "policies": [
            "No sensitive data exposure",
            "Professional tone"
          ]
        }
      },
      "observation": {
        "type": "tool_result",
        "output": {
          "version": "pvs-1",
          "approved": true,
          "reasoning": "Report contains appropriate technical details without exposing sensitive information.",
          "policy_violations": [],
          "confidence_score": 0.95
        }
      },
      "reflection": {
        "thought": "My report passed policy review. I can now deliver the final output.",
        "next_action_hint": "Deliver final report",
        "uncertainty": 0.05
      },
      "metadata": {
        "kind": "policy_verdict",
        "policy_engine": "the-gavel",
        "latency_ms": 450
      }
    }
  ],
  "final_output": {
    "type": "report",
    "content": "# Security Review Report\n\n## Summary\nReviewed `/src/auth/login.ts` and found 1 medium-severity issue.\n\n## Findings\n\n### 1. Input Validation (Medium)\n- **Location**: Line 42\n- **Issue**: User input not sanitized before database query\n- **Recommendation**: Use parameterized queries or input validation\n\n## Conclusion\nAddress the identified issue before deployment.",
    "format": "text/markdown"
  },
  "status": "succeeded",
  "error": null,
  "started_at": "2025-12-14T10:00:00.000Z",
  "completed_at": "2025-12-14T10:00:12.345Z",
  "metadata": {
    "total_tokens": 1319,
    "total_cost_usd": 0.0089,
    "models_used": [
      "gpt-4.1"
    ]
  }
}

Index Agent Data Protocol