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

Annotations only say readOnly=false, destructive=false, idempotent=false, openWorld=false; the description adds the behavior that matters — the agent must stop writing in the conversation and must disclose plainly to its own user but never to the other party. It still does not say what review actually does, whether the flag is reversible, or whether the other party is notified.

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 sentences, front-loaded with the action and effect, then the trigger list, then the exclusions, then the disclosure rule. The category enumeration is long but each item earns its place as a decision criterion.

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?

No output schema exists, so the description must cover the calling decision, and it does so thoroughly, including the post-call disclosure obligation. It is slightly incomplete on outcome (what review leads to, whether the agent can ever resume writing, whether the flag can be withdrawn), which an agent might need before committing.

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?

With schema coverage at 50%, 'reason' and 'messageIds' are already documented in the schema and the description adds nothing to them. It does implicitly map several of the prose triggers onto the category enum values (threat, sexual_minor, grooming, extortion, scam, self_harm, off_platform_pressure, harassment), which is useful, but the mapping is only inferable and conversationId remains undocumented anywhere.

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 ('Flag a conversation for human review') plus an immediate side effect ('stop writing in it'), which distinguishes it from read_chat, reply_in_thread, and block_person without opening any schema.

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 inclusion list (threats, grooming, extortion, scams, self-harm risk, off-platform pressure, repeated contact after a no) and an explicit exclusion list (disagreement, bluntness, one rude message, someone being different). The only gap is that it never names the nearest alternative (block_person) for the 'repeated contact' case.

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