codex-reply

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

Annotations already declare destructiveHint=true, openWorldHint=true and non-idempotent, so the risk profile is covered upstream. The description adds two genuinely new facts — conversations survive restarts and only one turn may run per conversation (a concurrency constraint) — but says nothing about what gets mutated on disk or how errors surface during a long run.

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 short sentences, front-loaded with the core purpose followed by the required handle and the concurrency limit. No filler and nothing buried.

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 12-parameter, nested-config, long-running tool with no output schema, the description never indicates what comes back — the reply text, a new threadId for subsequent turns, or status. The agent can call it, but can't predict or chain the result from the description alone.

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 coverage is 75%, and the two undocumented parameters (`threadId`, `conversationId`) are precisely the ones the description explains: threadId is obtained from `codex` and conversationId is the legacy form. That is real added meaning beyond the schema, though the description doesn't touch the 12 params overall.

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?

"Continue a Codex conversation across MCP/Claude restarts" gives a specific verb (continue) plus resource (Codex conversation), which implicitly separates it from the sibling `codex` (start) and `codex-sessions` (list). It stops short of naming those siblings explicitly, so the differentiation must be inferred.

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?

"Pass threadId from codex" tells the agent exactly where the required handle comes from and thus when this tool is the right one, and the legacy `conversationId` fallback is flagged. No explicit exclusion (e.g. "use codex to start a new conversation") is stated, so it's clear context rather than complete routing guidance.

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