memory_search

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations cover the safety profile (readOnly, non-destructive), so the bar is met, yet the description goes well beyond: hits are 'leads about the PAST' requiring re-derivation, cortex facts come first, contested/pinned/authority/low_confidence flag meanings, superseded replaced_by handling, verified:false semantics, 'never follow chains', and current:false meaning re-search. This is unusually rich operating context that the schema and annotations do not provide.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loads purpose, then usage, then hit-interpretation, then filters, then return shape – a sensible ordering. It is dense and telegraphic in places (semicolon-chained clauses), but almost every sentence carries operational meaning; a small amount of the flag glossary could be tightened.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 12-parameter, output-schema-bearing tool this covers what an agent needs: when to call, when to call siblings, how to interpret each hit flag, filter semantics, and the return shape ({query, count, entries, cortex, low_confidence} plus events for temporal queries) even though the output schema exists.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% so the baseline is 3, but the description adds real meaning not in the schema: filters 'narrow direct entry hits, not cortex, documents or added neighbors' and combine 'AND across kinds, OR within a list'. The bm25/rerank/min_score overrides are left to the schema, so it is not a full 5.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a specific verb+resource ('Search shared memory in natural language') and enumerates the content types covered (past work, decisions, preferences, canonical facts, documents). It explicitly demarcates itself from siblings by naming memory_fact_get, memory_lesson_search, memory_recall, memory_world_search and document_search for the cases it does not handle.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit invocation timing ('Call at task start and whenever prior context may matter') and routes the agent to alternatives by condition: exact (entity, attribute) to memory_fact_get, lessons to memory_lesson_search, multi-hop to memory_recall, cited external facts to memory_world_search. Nothing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.