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

The description goes well beyond the annotations. It warns that this is the tool that spends money, explains the held status with projected_bytes/projected_runtime_seconds/projected_cost, and clarifies that a sample must state a ceiling. The annotations already declare idempotentHint=true and destructiveHint=false, and the description does not contradict them; it adds the cost and hold behavior that annotations cannot express.

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 dense but every sentence earns its place: the cost warning is front-loaded, the mode definitions are compact, the example is illustrative, and the follow-up tool routing is at the end. It packs a lot of critical information into a few sentences without redundancy.

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 7-parameter tool with no output schema, the description is remarkably complete. It covers the return shape (run_id, status, mode, version, dashboard_url), the held status behavior, the mode-specific constraints, and the next steps in the workflow. The only minor gap is the lack of explicit mention of resource_class, but the description still provides enough for an agent to call the tool correctly in the common paths.

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 the meaning of mode (four values with one-line definitions), the window object (start/end), and the ceiling parameters (max_rows, max_source_bytes, window) in the context of sample and backfill. It does not explain resource_class or recipe_digest, but the example and the mode semantics cover the most decision-critical parameters. Given the 0% coverage, this is strong compensation, though not exhaustive.

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 ('Runs a registered recipe') and immediately distinguishes the four modes (sample, full, refresh, backfill) with one-line definitions. It also names the sibling tools it is not (confirm_run, run_events, query_run), so an agent can tell it apart from the surrounding run lifecycle tools without opening their schemas.

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 description gives explicit when-to-use guidance: 'always start here' for sample mode, and it states the exact conditions for sample and backfill (sample must state a ceiling; backfill must state a window). It also tells the agent what to do when a run is held (show projected costs and call confirm_run only if the user agrees), and it names the next tools to use (run_events, query_run). This is exemplary routing and usage guidance.

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