Start Codex task

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

The description adds behavioral context beyond annotations: it returns threadId, turnId, and an event cursor, and clarifies that the turn begins immediately in an allowed local repository. Annotations already mark it as destructive and open-world, so no contradiction exists. It doesn't elaborate on approval or sandbox behavior, but those are covered by schema and 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 concise sentences convey purpose, behavior, and return value without repetition. Every clause earns its place: it names the resource, the immediate action, the allowed scope, and the downstream tools that consume the return values.

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 no output schema, the description appropriately explains the return values. It also contextualizes the tool as a starting point for a thread lifecycle. It doesn't explicitly state that execution is asynchronous or that the returned IDs are needed for later calls, but those are strongly implied by the mention of the event cursor. The tool's complexity is well handled overall.

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 schema documents all parameters well. The description only adds 'allowed local repository' context that aligns with cwd, and mentions the return cursor, but doesn't add new details about parameters like model, effort, sandboxMode, etc. This meets the baseline for high schema coverage.

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 clearly states the action ('Start a new Codex app-server thread'), the resource ('Codex app-server thread'), and the immediate behavior ('begin a turn in an allowed local repository'). It also distinguishes this from sibling tools by explicitly returning an event cursor for codex_wait or codex_status, signaling this is the entry point for new threads.

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 explains what this tool does and implies its role relative to siblings by referencing post-start tools (codex_wait, codex_status). It doesn't explicitly say when not to use it or name alternatives for existing threads, but the context of 'new' thread is clear. The return cursor guidance helps the agent know how to follow up.

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