Single tool 'ewcorr_scan' has a basic description (89 chars) but critically lacks parameter descriptions. The 'target' parameter has only a generic type annotation with no guidance on expected format, valid values, or examples. No output schema is documented, the tool returns JSON but the structure is not formally specified. The description mentions 'Returns JSON findings' but does not explain what fields are in that JSON or when to use this tool vs alternatives. No error handling guidance. Naming is acceptable (verb_noun form) but the tool's purpose is domain-specific (EW correlation) without context for general LLM selection.

• Add a detailed description to the 'target' parameter: 'Path to an EW/ELINT observation log file or directory. Accepts absolute or relative paths. If a directory, recursively scans all files. Example: /var/logs/ew_events.log or ./observations/'
• Document the output schema in the tool description: 'Returns a JSON object with fields: {tool: string, findings: [{id, sev, where}], score: number}. See examples in README.'
• Expand tool description to 150-250 chars: 'Correlate electronic-warfare event logs by time/frequency/bearing to cluster emitters into candidate signal sources. Input: path to observation log file or directory. Output: JSON with emitter clusters, confidence scores, and sensor hits. Use this to identify and track passive EW targets.'
• Add error handling guidance: 'If the path does not exist, returns error with available directories. If the file format is invalid, returns partial results with a warning. Retry with a different path or check file permissions.'
• Add parameter constraints: 'target must be a valid filesystem path (string, 1-1024 chars). Relative paths are resolved from the current working directory.'