Codex Agents

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

The description discloses important stateful behavior beyond the annotations: pools survive MCP requests, route back to their bound workspace by id, and reconnects can reset implicit workspace selection. This is exactly the kind of contextual, non-obvious behavior an agent needs to know, and it does not contradict the provided 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?

Four dense sentences with the most critical warning front-loaded before the lifecycle advice. There is no filler, though 'stable h5i boxes' is jargon that could be clearer for an unfamiliar agent.

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 multi-action tool with no output schema, the description explains the persistence and reconnect model well but leaves the semantics of actions like wait, status, messages, checkpoint, and stop to be inferred from the enum and parameter names. It is adequate for tool selection but not fully complete for invoking every action correctly.

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 67%, and the description mostly re-emphasizes the documented root requirement for action=start rather than explaining undocumented parameters like force, agents, or advance_baseline. It adds lifecycle context but does not materially clarify individual parameter behavior beyond what the schema already provides.

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 identifies the tool as a 'unified persistent codex agent interface' and clearly describes the start/send lifecycle with 'Start once, then send many tasks', so an agent can tell it governs persistent Codex agents. It does not explicitly contrast it with sibling tools like bash or workspace, so it falls short of a 5.

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 gives a concrete, high-value usage rule: for action=start ALWAYS pass the explicit Git repository root because MCP reconnects can reset implicit workspace selection. It also recommends a lifecycle pattern ('Start once, then send many tasks'), but it does not explicitly state when not to use the tool or route to an alternative.

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