Single tool 'dpiaforge_scan' has a basic description (58 chars) but critically lacks parameter descriptions. The input schema defines 'target' as a string with a description ('Target path or activity identifier to scan'), which is present but vague. No output schema is documented, the tool returns JSON findings but the structure is not formally specified. The tool name uses a verb prefix (scan) which is good, but the overall definition is sparse and would not guide an LLM effectively on when/how to use it. No error handling guidance, no pagination, no constraints on the 'target' parameter (could be a path, activity ID, or something else, ambiguous).

• Document the complete output schema: e.g., {"tool": string, "findings": [{"id": string, "sev": string, "where": string}], "score": number}. This lets the LLM understand what to extract.
• Expand the tool description to 100-200 characters. Explain: (1) What it does (scans for compliance issues), (2) When to use it (before deploying AI systems), (3) What it returns (JSON findings with severity levels).
• Add constraints to the 'target' parameter: 'A file path (absolute or relative) or directory to scan. Must be readable by the server. E.g., "/app/src" or "./config".'
• Add error handling guidance: 'If target is not found, returns error with available paths. If permission denied, returns error with required permissions.'
• Consider adding optional parameters: 'limit' (max findings to return, default 100), 'severity_filter' (enum: high|medium|low|all, default all) to support filtering and pagination.