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

Annotations already declare readOnly, openWorld, and idempotent hints, so the description doesn't need to restate safety. It goes beyond annotations by disclosing the return structure (source, citation_uri, raw payload) and the stateful mark_read behavior that affects subsequent calls. This adds meaningful behavioral context not present in the schema or 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?

The description is four sentences, front-loaded with the core purpose, and every sentence adds value: purpose, return payload, filtering options, stateful behavior, and alternative access. There is no redundancy or filler, making it efficient and easy to parse.

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?

Given there is no output schema, the description compensates by describing the return payload fields and behavior. It also covers filtering, mark_read, and polling guidance, which are essential for an agent to use the tool effectively. It doesn't mention limit or unread_only, but those are fully documented in the input schema, so the description covers the remaining context adequately for a list tool.

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 100%, so baseline applies. The description adds extra meaning by providing a concrete type example ('sec_8k'), clarifying the since parameter expects an ISO timestamp, and explaining how mark_read affects next calls. These details go beyond the schema's property descriptions and help agents use the parameters correctly.

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 specific verb and resource: 'Pull fired events from your subscription feed.' It clarifies the tool returns recent alerts from the evaluator's persisted feed, including source, citation_uri, and raw event payload. This clearly distinguishes it from sibling tools like list_subscriptions or recent_changes, which handle different resources.

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 practical guidance: filtering by type/since, using mark_read to advance the read cursor, and note that 'Polls work fine' to indicate suitability for repeated calls. It also mentions the same feed is available via an HTTP endpoint for scripts/dashboards, providing an explicit alternative. It doesn't compare directly to sibling tools but offers clear context.

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