Critical gaps across all four tools. No input schemas are visible in the source code, only tool names and generic descriptions are provided. The descriptions are extremely brief (under 100 chars each) and lack actionable detail about parameters, expected inputs, outputs, or error handling. Tools are registered via `quantakryptoTools` from an external package (@quantakrypto/mcp), but their actual schemas, parameter definitions, and output structures are not visible in this codebase. Per the hard rules, tools without visible schemas score 0 on schema, and tools with descriptions under 20 chars score 0-20 on description. The gateway filters tools but does not define or document them.

• Export or re-document tool schemas from @quantakrypto/mcp in this gateway codebase so they are visible for evaluation and can be audited for LLM-friendliness.
• Expand each tool description to 100 - 200 characters, explaining: what the tool does, when to use it (vs similar tools), what inputs it expects, and what it returns. Example: 'Analyzes submitted code to verify a security fix is correct. Requires code_snippet (string) and expected_fix_type (enum: crypto_vulnerability|key_exposure|algorithm_weakness). Returns verdict (pass|fail) and explanation.'
• Document output schemas for each tool, including field names, types, and what each field means. This enables LLMs to chain tools and extract required data.
• Add error handling guidance: document common failure modes (invalid input, timeout, service unavailable) and what the LLM should do (retry, ask user, try alternative tool).
• If tools are defined in @quantakrypto/mcp, consider adding a local wrapper or adapter layer in this gateway that re-exports and documents them with LLM-optimized descriptions and schemas.