claude_request_async

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

Annotations already carry the safety profile (readOnlyHint=false, destructiveHint=true, idempotentHint=false, openWorldHint=true), so the description's burden is reduced. It adds genuinely useful context: 'durable' implies the job outlives the invocation, and the poll/collect workflow clarifies the lifecycle. However, it does not disclose that launching a job consumes tokens/cost or may execute arbitrary Claude Code operations beyond what the destructiveHint already flags. No contradiction with 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?

Two sentences, zero filler. The primary action is front-loaded, the durability qualifier is essential, and the second sentence gives the complete continuation workflow. Every word earns its place.

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 tool with 52 parameters, nested objects, and no output schema, the description is quite thin: it never states what the tool returns (presumably a job identifier to pass to llm_job_status/llm_job_result), and it doesn't touch on cost, cleanup, or session persistence. The exhaustive 100%-coverage schema and the rich annotation set compensate heavily, but the missing return contract is a real gap for an agent deciding how to chain calls.

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 the baseline of 3 applies even though the tool description itself adds zero parameter information. The schema's parameter descriptions are unusually detailed (flag spellings, approval gates, provider restrictions, mutual exclusions), so the description does not need to compensate. There is no gap to penalize beyond the baseline.

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 ('Start'), a specific resource ('a Claude Code CLI request'), and the distinguishing mode ('durable background job'). This clearly differentiates it from the synchronous claude_request sibling and the other *_request_async variants, while 'Poll with llm_job_status, collect with llm_job_result' names the exact follow-up tools. An agent can understand what this does and how it differs from siblings without opening any schema.

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 phrase 'durable background job' conveys the core selection criterion (async vs. the sync sibling), and naming llm_job_status/llm_job_result gives the agent the polling/collection workflow. However, it never explicitly says 'use claude_request when you need an inline synchronous result' or states when NOT to use this variant, leaving the sync/async decision partly implicit.

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