memory_reconcile_apply

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

With no annotations provided, the description carries the full burden and does so richly: it documents Zep-style invalidation semantics (merge sets memo_ids invalid with invalidated_by rather than deleting), per-op error isolation, idempotency on already-invalidated items, and the enforce red-line quotas (≤400 chars per entry, global 1200 / project 1500 / user 300) with over-limit errors returning usage. That is exactly the behavioral context an agent needs before invoking.

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?

The purpose and determinism/idempotency claim are front-loaded, then each op type is a compact bullet with its fields. It is dense but every bullet maps to a distinct dispatch branch, so the length is largely justified; a small amount of explanatory padding (the code-path reference) could be trimmed.

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?

An output schema exists, so return-value explanation is unnecessary, and the description fills the remaining gaps: per-op semantics, mixed-op batching, per-item error isolation, red-line constraints, and the side effect of refreshing last_reconcile_at. Nothing an agent needs to invoke it correctly appears to be missing.

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%, but the schema only declares an array of unconstrained objects with an 'op' discriminator. The description goes well beyond the schema by fully specifying the item shapes for merge, invalidate, score, promote (including kind enum values and scope options) and keep/noop, which is essential disambiguation the schema cannot provide.

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?

The description states a specific verb+resource: batch-applying LLM-adjudicated memory reconciliation operations, with explicit determinism and idempotency. It distinguishes itself contextually from the candidate-generation flow ('LLM 精判确认后'), but it names no sibling tool directly, so the differentiation is implied rather than stated.

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?

It clearly positions the tool: use it to apply operations after LLM adjudication has confirmed them, and it explains that keep/noop can be omitted and that unknown-field/unknown-op cases return an error without blocking. It does not explicitly name an alternative (e.g. memory_reconcile_candidates or memory_invalidate) or state when not to use it.

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