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

Annotations declare readOnlyHint, openWorldHint=false, and idempotentHint. The description goes well beyond these by explaining that sentiment is written once per lead and null means unscored, that unscored replies are missing from get_campaign_stats, that derived fields are structural and can be fooled (e.g., a broadcast sent to one lead), and that Onsa cannot see manual LinkedIn replies. No contradictions; the description adds rich behavioral context.

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?

The description is lengthy but densely packed with essential nuances, structured with bolded field names and clear paragraphs. It is front-loaded with the core purpose and then elaborates on derived fields and pitfalls. While it is verbose, every sentence carries operational value; no filler exists. It could be slightly trimmed, but the structure and labeling make it highly scannable.

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?

Despite having an output schema, the description explains the semantics of derived fields (awaitingOurReply, daysSinceLastReply, looksLikeBroadcast) and their interpretation, including edge cases. It clarifies how the tool relates to get_campaign_stats and the funnel, and gives concrete guidance on handling null sentiment and missing manual replies. For a tool with such nuanced data, the description is exceptionally 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?

The input schema covers both parameters (campaignId and limit) with descriptions, giving 100% coverage, so the baseline is 3. The description adds a practical note about raising limit if awaitingOurReplyCount exceeds returned, which is a useful semantic hint beyond the schema. It doesn't describe parameter formats, but the schema already does. This slight added guidance justifies a 4.

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 opens with a clear verb and resource: 'THE ACTUAL WORDS people replied with, for every lead in this campaign that answered.' It immediately differentiates from get_campaign_stats by contrasting counts/labels with the actual text, and states the only way to understand cohort reactions. This makes the tool's purpose unmistakable and distinct from its siblings.

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?

The description gives explicit when-to-use guidance: 'Use it before you draw any conclusion about an ICP from response counts.' It also contrasts with get_campaign_stats and provides nuanced instructions on interpreting derived fields, such as not saying 'unanswered' unless awaitingOurReply is true, and not counting broadcast-looking messages as engagement. It also cautions against over-trusting labels, guiding the agent on when to use this tool vs. relying on stats.

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