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

Beyond annotations, the description discloses that the tool reads sources, can run for a long time, uses one progressive acquisition with an automatic inspection checkpoint after about five minutes, and can return status 'held' with projected_bytes and projected_runtime_seconds. This prevents the agent from misinterpreting the checkpoint as an approval gate or assuming immediate completion.

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 long but dense with operational detail; every sentence contributes a distinct fact or guardrail, including a JSON example, return shape, and held-state behavior. It is front-loaded with the core purpose before adding workflow instructions.

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, no-output-schema tool with only sparse annotations, the description covers invocation modes, constraints, return fields, held behavior, and next-step tools. The main omission is resource_class semantics, and window format is left to the schema, so it is highly complete but not exhaustive.

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 description coverage, the description compensates by explaining the mode enum, requiring a ceiling for sample mode and a window for backfill mode, and showing an example with recipe_id, recipe_digest, and mode. However, resource_class is not explained, and the exact format of window start/end is left to the schema, so not all seven parameters are fully covered.

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?

Description opens with 'Runs a registered recipe' and reinforces 'THIS IS THE TOOL THAT DOES THE REAL WORK', giving a specific verb and resource. It also distinguishes start_run from confirm_run and cancel_run by clarifying the run lifecycle and that the checkpoint is not an approval gate.

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 prescribes 'Start builds in full mode' and defines all four modes with field-level constraints: a sample needs at least one ceiling and a backfill needs a window. It also gives the post-call workflow ('Next: run_events... then query_run') and conditionally routes to confirm_run only when a held run's projections are accepted.

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