agy_project_update

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

Goes well beyond the annotations: discloses size ceilings (memory <=4500 UTF-16 units, plan <=4000), that an empty plan clears it, lifecycle semantics (pausing blocks NEW calls but does not cancel active ones; completed projects need an explicit active update), no checkpoint expiry, and that no automatic retries/model calls/credential migration occur. This is exactly the non-obvious behavioral context annotations cannot carry.

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 create/update decision and every sentence carries substantive facts; no filler. However, it is a dense semicolon-chained block with mixed concerns (schema shape, lifecycle, limits, retries) that could be segmented for faster scanning.

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 10-parameter mutation with no output schema, the description covers concurrency (expected_version), payload shape, size limits, lifecycle side effects, and failure/retry behavior. It does not say what the update returns (e.g., new version) or address the required reason field, a minor remaining gap.

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?

With 0% schema coverage across 10 parameters, the description compensates heavily: it explains expected_version=0 creation semantics, the memory sub-shape (facts/disagreements/next_steps with {text,sources} and positions {participant,view,reason,sources}), the source-string format discussion_id#message_id, plan_sources being required with plan, and status values with their effects. The required 'reason' parameter and the null-vs-omitted distinction are left unexplained.

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?

States a specific verb+resource ('Create/update a lasting project checkpoint') and immediately splits the two modes (create via expected_version=0 vs update after reading current version). An agent can tell this apart from sibling tools like agy_project_read or agy_project_link without opening the 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?

Gives clear conditional guidance: create with expected_version=0 plus name and goal, otherwise read the current version first; also explains that capture_discussion_id is a retry path for local sync and 'never models'. It stops short of explicitly naming the read sibling (agy_project_read), leaving the alternative implicit.

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