delegate

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

The description discloses mode-specific behavior (agent edits, plan returns structured plan, ask is read-only, review runs native review) and parameter constraints, going well beyond the annotations' readOnly/destructive hints. It also clarifies the 'read-only' nature of ask mode without contradicting the tool-level readOnlyHint false, since agent mode performs edits.

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?

Four compact sentences cover purpose, usage guardrail, mode semantics, and parameter constraints. Every sentence adds essential information, and the most important directive ('Never shell out to codex') appears immediately after the purpose.

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 10 params, 4 modes, and an output schema, the description is succinct yet comprehensive. It covers the core use, mode behaviors, parameter discipline, and points to additional skill documentation. With an output schema present, the lack of return-value details is acceptable.

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?

The schema already describes most parameters (90% coverage), but the description adds critical semantics for mode values, explains when to touch model/reasoningEffort/fast, and clarifies the 'ask' mode is read-only. This provides value beyond the structured schema fields.

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 'Delegate a coding task to the OpenAI Codex CLI' with a specific verb and resource. It distinguishes itself from sibling tools (cancel, doctor) by being the main delegation entry point, and explicitly frames the tool as the only sanctioned way to invoke Codex ('Never shell out to codex — use this tool').

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?

Provides explicit when-to-use guidance: 'Never shell out to codex — use this tool' and explains each mode's purpose. It also instructs when NOT to change certain parameters ('only when the user asks') and points to additional orchestration details via the delegate skill.

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