PFAA Engine exposes 37 tools via STDIO with minimal schema documentation. Most tools lack formal input schemas in the source code provided, only parameter names and types are inferred from descriptions. No output schemas are documented. Descriptions are present but generic (avg ~60 chars, well below the 194-char baseline). Parameter descriptions exist but lack constraints, ranges, or format guidance. No error handling guidance, no recovery paths, and no tool annotations (readOnlyHint/destructiveHint). The server is a custom Python implementation without visible MCP protocol compliance. This is a typical community-grade tool collection, not production-ready.

• Add explicit JSON Schema definitions for all 37 tools with type, description, and constraints for every parameter. Use enums for voice names, file_glob patterns, and status values.
• Expand tool descriptions to 100-200 chars. Include WHEN to use, prerequisites, and side-effect warnings. E.g., 'shell: Execute a shell command. WARNING: Destructive commands (rm, dd) cannot be undone. Use only for safe operations like ls, grep, echo.'
• Add readOnlyHint, destructiveHint, and idempotentHint annotations to all tools. Mark shell, trash_file, git_branch, volume_set, dark_mode as destructive.
• Document output schemas for all tools. Specify return types (string, object, array), field names, and structure. E.g., git_log returns array of {commit_hash, author, message, date}.
• Add error handling guidance. E.g., 'If shell returns exit code 127, command not found, check PATH. If browser_check times out, the URL may be unreachable or slow, retry with longer wait_ms.'