Server provides 7 tools with complete input schemas and descriptions. Naming follows verb_noun convention (actinf_*). However, descriptions lack depth on WHEN/WHY to use each tool and lack recovery guidance for errors. Parameter descriptions are present but generic. Output schemas are not documented, LLMs cannot predict response structure. No error handling guidance visible. Tool composition is reasonable but lacks idempotency guarantees and chaining metadata.

• Document output schema for each tool. Example: actinf_get_state should return {regime: string, belief_distribution: object, uncertainty: number, goal_tracking: object, loop_status: string}. This enables LLMs to chain tools and extract required fields.
• Expand descriptions to 100-200 chars with WHEN/WHY guidance. E.g. 'Get current cognitive regime, belief distribution, uncertainty, goal tracking, and loop status. Call this before prescribe_policy to understand the current state and decide if a policy update is needed.'
• Add error handling guidance to actinf_evaluate_action: 'Returns {verdict: ALLOW|MODIFY|DENY|ASK, confidence: number, reason: string}. If DENY, the proposed action violates safety constraints, ask the user for confirmation or suggest an alternative action.'
• Constrain context_attributes with a schema or enum. Document valid keys (failure_signal, risk, context_pressure) and expected value types. Replace free-form string with structured object parameter.
• Declare idempotency for state-modifying tools. Add to descriptions: 'Idempotent: calling with the same input multiple times produces the same result without duplicate side effects.' This guides agent retry logic.