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

With annotations declaring readOnlyHint=false and idempotentHint=false, the description explains exactly why: "Recording this delivery is what makes the next recall's change exact" — i.e., the call mutates delivery state and therefore is not a pure read and not repeatable with identical results. It also discloses the auth requirement ("Needs your key") and anchors the payload to a known endpoint, which is high-value context beyond the annotations.

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?

Dense and mostly front-loaded around the core recall behavior, with every clause carrying information about state effects or auth. The leading "Watches: configure, unfurl and save free" is the weakest sentence — telegraphic and slightly confusing rather than routing-clear.

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 stateful, non-idempotent recall with a small schema and no output schema, the description covers the essential unknowns: it is a write-ish delivery record, it requires a key, and it returns a current reading plus a delta. The only soft spot is that the return shape is referenced via an external API path and version string rather than described inline.

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 both watchId and expand are already documented in the schema; baseline 3 applies. The description reinforces the semantic link between the call and the diff window ("since the last time it was delivered"), but adds nothing about the expand depth override.

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?

"Recall a saved Watch: its reading now (sb.watch-reading.v1) and what changed since the last time it was delivered to you" gives a specific verb (recall) and resource (saved Watch) plus the two things returned. The opening fragment "Watches: configure, unfurl and save free" gestures at sibling capabilities (unfurl_watch, save_watch) but is phrased ambiguously enough that the differentiation is only implied.

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?

The phrase "configure, unfurl and save free" loosely signals that those operations belong to other tools, and "recall a saved Watch" implies a watch must already be saved. However, there is no explicit "use X instead" routing and no stated precondition (e.g., that watchId must come from list_saved_watches or save_watch).

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