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

The description goes well beyond the annotations by disclosing concrete consequences: it actually submits inside the employer's system, applications leave in the person's name, it never re-applies to jobs already applied to, and it consumes plan quota. This is high-stakes behavioral context that an agent needs before invoking the tool, and it aligns with destructiveHint, openWorldHint, and idempotentHint.

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 action, the safety valve, the prerequisites, the side effects, and the follow-up polling step. The most important consequence—real submission—is front-loaded, and the dry-run escape hatch appears early enough to prevent accidental real applications.

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, destructive, externally visible operation with no output schema, the description is remarkably complete: prerequisites, side effects, plan usage, filter semantics, duplicate avoidance, dry-run mode, and status polling are all covered. An agent has enough information to decide whether to call it, how to call it safely, and what to do afterward.

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 only 50% schema description coverage, the description compensates by explaining `limit` (takes up to limit matching jobs) and `dry_run` (does everything except submit). The `fresh` and `match` parameters remain schema-covered with defaults explained, and tying them to 'same filters as list_matching_jobs' adds valuable cross-tool context.

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 states a specific verb and resource: 'Start an application run' and then explains exactly what that entails—tailoring CV/cover letters, answering screening questions, and submitting inside the employer's system. It clearly differentiates from siblings by referencing list_matching_jobs filters and get_apply_run_status for polling, so an agent can distinguish it from preview, listing, or status tools.

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 strong contextual usage signals: real applications are submitted, prerequisites are a confirmed email and active plan, it counts against the plan, and dry_run=true is the non-submitting safe path. It does not explicitly name alternatives like preview_tailored_cv or stop_apply_run, but it clearly implies when the tool is appropriate and what preconditions must hold.

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