This is a stub/reference implementation with severe definition gaps. All 5 tools have basic parameter schemas visible in the Rust code, but descriptions are minimal (10-30 chars), lack actionable context, and do not explain when/why to use each tool. No output schemas are documented. The 'exec' tool is particularly dangerous, it accepts arbitrary commands with no validation guidance, no dry-run option, and no error recovery hints. Parameter descriptions are generic ('The command and arguments to execute') without format constraints or examples of valid inputs. No tool explains dependencies, prerequisites, or what data structures are returned. This server is not production-ready.

• Expand tool descriptions to 100 - 200 characters. For 'exec': 'Validate and execute a shell command against the active OATP toolset. Checks policies (allow/deny/forbidden_args), phase gates, and preconditions before execution. Returns exit code, stdout, stderr. Irreversible, no dry-run available; use check tool to validate toolsets.json first.'
• Add enum constraints to 'phase' parameter: {"type": "string", "enum": ["reconnaissance", "surgery", "instrumentation"], "description": "Phase to activate. reconnaissance: discovery and analysis. surgery: invasive modifications. instrumentation: monitoring and tracing."}
• Document output schemas for each tool. E.g., 'ls' returns {"tools": [{"name": string, "phase": string, "category": string, "required": boolean, "verification": string}], "total": number}.
• Add dry-run or confirmation support to 'exec'. E.g., add optional 'dry_run' boolean parameter: 'If true, validate the command and return what would execute without running it.'
• Add error recovery hints to descriptions. E.g., 'check' tool: 'Validates a toolsets.json file. On schema error, returns the invalid field and constraint violated. Use this before exec to catch configuration issues early.'