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

With readOnlyHint=false and openWorldHint=true, the annotations only say 'may have effects and reaches outside systems' — the description supplies the critical missing mechanic: anything external is returned as a pending action with an id and nothing is emitted until confirm_herman_action. That two-step staging behavior is exactly the context an agent needs to avoid assuming a real-world send already happened. It also reconciles the apparent tension between a side-effect-free feel and the write-ish 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?

Two sentences, front-loaded with the actor and its scope, then immediately the staging rule and the follow-up tool. No filler and no restatement of the title or schema.

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 single-parameter natural-language router with no output schema, the description covers purpose, the staged-result shape ('pending action with an id'), and the required next step. What is left out is failure/ambiguity handling — e.g. what happens if Herman cannot map the request or whether the reply includes clarifying questions — which would matter for a free-text entry point. Overall it is nearly complete.

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 100% and there is a single required string, so the baseline is 3. The description adds beyond the schema's 'What you want, in plain words' by giving concrete example payloads (status, finding companies, running campaigns, sending the crew to call), which shows the expected granularity of a natural-language request. It does not discuss ambiguity resolution or follow-up clarification, so not a 5.

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 names a specific resource — Herman, the manager agent that runs the bullpen — and gives a concrete capability list (status, finding companies, running campaigns, sending the crew to call). This makes it distinguishable from direct sibling tools like create_campaign, find_companies, or send_email, which cover one slice each while this is the natural-language orchestrator. It stops short of an explicit 'this is the router, prefer it over calling the individual tools' statement.

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?

Usage is conveyed through the enumerated domains an agent can ask about (status, prospecting, campaigns, calling), which implicitly tells the caller when this tool applies. It does not spell out exclusions or explicitly route an agent between ask_herman and the direct action tools when both could serve. The handoff to send-confirmation is clear.

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