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

The annotations only declare the safety profile (not read-only, idempotent, not destructive); the description adds substantial context beyond them: the idempotent-replay behavior costs no credits, successful calls cost 1 entry credit while failures are free, superseded entries are retired with a pointer, and bodies over 600 characters are echoed back as `body_truncated`. It also discloses an edge case — a newly created ledger triggering `new_project`/`known_projects`/`hint` in the response — which no structured field conveys.

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-loaded with the core instruction, then parameters, then return shape, then cost — a sensible order that lets the agent stop reading early. It is dense and every sentence carries information, but it is long enough that the return-shape and new-ledger details could be tightened slightly.

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?

With no output schema, the description carries the return contract itself and does so: entry id format (KNOW-n/TASK-n), truncation signaling, replay returns the existing entry, and the new-ledger result fields. Combined with cost and supersedes semantics, an agent has everything needed to call this correctly and interpret the result.

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 on top: it explains what belongs in `body` (what happened, why, what to do instead), that `agent` is for your name while the server separately records the key, and that `supersedes` retires the referenced entry rather than merely linking. The one gap is that `project` naming semantics are left to the schema.

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 opens with a concrete verb and resource ("Write down something... a decision, lesson, guardrail, note, or task") and enumerates the exact content types so the agent knows precisely what belongs here. It is clearly distinguishable from the listed siblings, which read (recall), delete (retire_entry), summarize (brief), and assign tasks (claim_task).

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 gives explicit triggering conditions: record a lesson when a change is rolled back or a fix took digging, a decision when an approach is chosen over another. It also directs the replace-an-earlier-entry case to `supersedes`. What's missing is any exclusion or cross-reference — nothing tells the agent to check recall for duplicates first, or how this differs from retire_entry when retiring old data.

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