Single tool 'deeplinkfuzz_scan' has a reasonable description (127 chars, within 10-1024 range) but critical schema deficiencies. Input schema is minimal: only 'target' parameter with type 'string' and basic description. No output schema documented. Parameter lacks format constraints, range limits, or validation guidance. No error handling patterns, no recovery guidance, no idempotency declaration. Tool name uses verb_noun pattern correctly but lacks composition clarity, unclear what 'fuzzes' returns or how to chain results. No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Description does not explain prerequisites (emulator/device availability), expected input format (path vs manifest file), or output structure. Baseline: 100% of A+ tools document return types; this tool does not.

• Document the complete output schema: define the structure of the JSON response (e.g., {findings: [{id, severity, component, payload}], score: number, tool: string}). Include field types and descriptions.
• Add format and validation constraints to 'target' parameter: specify accepted formats (e.g., 'Path to AndroidManifest.xml file or directory containing manifests'), expected file extensions, and behavior for missing/invalid paths.
• Add tool annotations to the FastMCP decorator: @app.tool(readOnlyHint=True, idempotentHint=True) to signal that this tool is safe to retry and does not modify state.
• Expand description to include prerequisites: 'Requires an Android/iOS emulator or device to be running and accessible. Returns JSON findings with id, severity, component, and payload fields.'
• Add error handling guidance: document common failure modes (emulator not running, invalid manifest format, permission denied) and suggest recovery steps (e.g., 'Start emulator with: adb start-server').