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

With zero annotations, the description carries the full burden and delivers: status defaults to 'open', the non-obvious side effect that adding an open task to a finished epic reverts it to 'designed', the VSCode config-file override for projectName, and the semantic rule that descriptions are summaries. This discloses exactly the behaviors an agent cannot infer from the schema.

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?

Roughly 150 words for a 13-parameter tool, and every sentence carries a distinct operational rule — no filler. The only deduction is structural: it is a single dense wall of text; short bullets or paragraph breaks would make the ~10 distinct instructions more scan-friendly for an agent, though nothing is redundant.

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 complex 13-param mutation with no annotations and no output schema, the description covers the action, defaults, edge-case state transitions, prohibitions, config override, linking strategy, and policy constraints — remarkably complete. The only gap is return-value expectations (e.g., does it echo the created task or its ID?), which would matter since no output schema exists.

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 only 46%, and the description compensates strongly: epicId gains omit-vs-provide semantics, status gains its default, projectName gains the config-file precedence rule, blockedReason gains a conditional requirement, priority gains 'leave unset if unknown', description gains 'summaries only', and documentId gains a spec-linking strategy. Seven of thirteen parameters receive meaning beyond the schema, well exceeding what a 3-point baseline would allow.

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?

Opens with a specific verb+resource ('Create a task') and immediately differentiates within a crowded create_* family (create_epic, create_document, create_decision) plus roster of task siblings (add_task_comment, update_task, list_tasks). Subsequent sentences clarify that it creates real tasks with epic linkage and status handling, so an agent cannot confuse it with create_epic or create_document.

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?

Exceptional routing guidance: 'omit for a standalone task', explicit exclusions ('Never create tasks for deploying/rolling out or for an acceptance checklist'), explicit redirection to alternatives ('use a standalone task or a new epic instead', 'create a document and link it with documentId'), and a pointer to a sibling for scope validation ('see list_tasks assignee="me"'). Also conditions the blocked status on providing blockedReason.

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