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

Annotations cover the safety core (readOnlyHint, destructiveHint=false, idempotentHint), but the description adds genuinely non-obvious behavior: the default acknowledge advances the caller's marker, while acknowledge=false peeks without moving it, and the result is caller-relative (which of YOUR reads are stale). That marker-advance side effect is important context not derivable from the annotations. It does not cover pagination or ordering of results.

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 core payload is front-loaded in one dense clause, followed by the acknowledge caveat and the sibling pointer. Nearly every clause carries information, though the payload enumeration is packed into a single run-on sentence that could be parsed more easily with light structuring.

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-value burden and does so well: checksum, document delta, staleness of the caller's reads, and per-revision who/when/why. The remaining gap is pagination behavior for result_offset and whether acknowledgement is per-project, which an agent would still need to discover.

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 description coverage is 0%, so the description must compensate. It explains acknowledge well (default advances the marker; false only looks) and ties the call to project_id, but result_offset — the pagination parameter — gets no mention at all, leaving one of three parameters entirely undocumented.

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?

It states a precise verb and resource — a delta read of what changed in a project since the agent's last index read — and enumerates the payload (checksum, added/removed/revised documents, stale reads, change provenance). It also explicitly contrasts itself with read_project, so an agent can pick between them without opening either schema.

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 clear context for when to call it (getting change deltas since the last index read) and one explicit alternative (read_project, which carries a short project_state). It does not, however, distinguish it from other change-oriented siblings such as list_changes, read_changes_summary, or read_attention, so sibling routing among the change tools remains partly inferential.

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