Loading ToolBench
Active Directory attack path mapper — min-cost paths + remediation priority
REDPATH exposes a single tool 'scan' via a wrapper (build_mcp_server) that abstracts the actual tool definition. The input schema visible in mcp_server.py shows only a high-level 'graph' object parameter with minimal structure. No explicit parameter descriptions, no output schema documentation, and no error handling guidance are visible in the source. The tool description is present but generic. The actual tool registration and schema details are hidden behind cognis_core.mcp.build_mcp_server(), making it impossible to verify schema completeness, parameter validation, or output structure from the provided code.
Active Directory attack path mapper — min-cost paths + remediation priority
Tool definition is abstracted behind build_mcp_server() wrapper, actual schema, parameter descriptions, and output structure are not visible in source code. Cannot verify compliance with JSON Schema requirements or parameter documentation standards.
Input schema shows only a bare 'graph' object with description 'Directed graph of AD principals and abuse edges with nodes, edges, owned principals, and targets', no nested field definitions, no type constraints, no validation rules visible. Does not meet JSON Schema completeness baseline.
No output schema documented. LLMs cannot infer what fields the 'scan' tool returns, preventing downstream tool chaining and forcing agents to guess at response structure.
No error handling guidance. Tool description does not explain failure modes, recovery steps, or what errors the LLM should expect (e.g., invalid graph format, missing required fields).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 35 | 2026-07-28+ | v2 |
Tool name 'scan' is generic and does not follow verb_noun convention. 'analyze_attack_paths' or 'compute_ad_paths' would be more specific and self-documenting for LLM tool selection.