CaseLaw MCP has 16 well-named tools with consistent verb_noun patterns (search_*, get_*, find_*). All tools have descriptions (avg ~80 chars, within baseline 34-392). Input schemas are present and typed for all tools. However, parameter descriptions are sparse or missing for many fields (e.g., 'court' in search_precedent lacks detail on valid values; 'display' and 'page' lack range constraints). Output schemas are not documented, LLMs cannot predict response structure. Error handling is absent from visible code. No tool annotations (readOnlyHint, destructiveHint). Descriptions are functional but terse; they lack WHEN-to-use guidance and dependency hints. Overall: solid foundation, but falls short of production-grade due to underdocumented parameters and missing output schemas.

• Document output schemas for all tools. For search_* tools, specify: {results: [{id, title, date, court/tribunal, ...}], total_count, page, has_next}. For get_* tools, specify: {id, title, body, references, ...}. Include field types and whether fields are always present.
• Add enum constraints to parameters with known valid values. E.g., search_precedent 'court' should list valid aliases; 'search_scope' should be enum ['title', 'body']. Replace description examples with formal constraints.
• Add min/max bounds to numeric parameters. E.g., 'display': {type: integer, minimum: 1, maximum: 100}, 'page': {type: integer, minimum: 1}. Update descriptions to state the range.
• Add error handling documentation to each tool. E.g., 'Returns {error: "not_found", message: "Precedent ID not found. Try search_precedent() first."}' for get_precedent. Include retryable vs fatal error guidance.
• Add tool annotations. Mark all search_* and get_* tools with readOnlyHint: true. This signals to agents that these tools are safe to retry and have no side effects.

Inferred effective spec: 2026-07-28+.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry