Start a new Codex thread

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

Annotations already declare destructiveHint=true, readOnlyHint=false, openWorldHint=true and idempotentHint=false, so the mutation/world-facing profile is covered structurally. The description adds real context beyond that: Desktop-mode atomic creation and assignment of a visible task, the prompt requirement implied by CODEX_BRIDGE_DESKTOP_TASKS=1, and retry semantics via a stable requestId. Minor tension with idempotentHint=false, since reusing requestId on retries is presented as safe, but this describes dedup identity rather than a true contradiction.

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?

Three dense sentences, front-loaded with the authorization condition and then routing; every clause carries decision-relevant information. It is slightly packed with mode-specific detail, but there is no filler.

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 5-parameter, non-idempotent, open-world creation tool with no output schema, the description covers eligibility, mode differences, retry identity and sibling routing. It never says what the call yields (e.g., a new threadId needed by send_to_codex_thread), which is the one gap for an agent that must 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 is 3; every parameter is already documented in the schema. The description reinforces the prompt/requestId relationship ('include the initial prompt and a fresh requestId ... keep requestId unchanged on retries') but adds no syntax, format, or constraint detail beyond what the schema already states.

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?

States a specific verb and resource — 'Start a new Codex task' — and immediately scopes it to authorized new conversations for independent work. It explicitly tells the agent how this differs from send_to_codex_thread (continuing existing work) and delegate_to_codex (waiting for a reply), so it is separable from siblings 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?

Gives an explicit precondition (user authorization, explicit or standing) and names two alternatives with the conditions that select them: 'Continue unfinished work with send_to_codex_thread and its original threadId' and 'Use delegate_to_codex to also wait for its reply.' Nothing is left to inference.

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