Single tool 'alertmux_scan' has a basic description (67 chars) but critically lacks parameter descriptions. The input schema defines 'target' as a string with a description ('Target path to scan for alerts'), which is minimal but present. However, the tool description does not explain what constitutes a valid target, expected output format, or error conditions. No output schema is documented. The tool name follows verb_noun convention (scan), but the description is generic and does not guide LLM selection or explain when to use it vs. alternatives. No error handling guidance, no pagination, no structured output documentation.

• Document output schema: 'Returns {tool: string, findings: [{id, sev, where}], score: number}'. Specify field types and meanings so LLM can extract and chain results.
• Expand parameter description: 'target: File system path (absolute or relative) to scan for code markers (TODO, FIXME, XXX). Must be readable directory or file. Returns findings with severity levels.'
• Rewrite tool description to clarify actual behavior: 'Scans files in a directory tree for code markers (TODO, FIXME, XXX) and returns findings with severity. Use this to identify technical debt and incomplete work in a codebase.'
• Add error handling guidance: 'If path not found, verify it exists and is readable. If no findings, target may not contain markers or may be empty.'
• Add constraints to parameter: 'target must be a valid file system path; max scan depth 1000 levels; returns up to 10,000 findings per call.'