CRM CSV Preflight (Sales List Rescue)

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

Annotations already cover read-only, idempotent, non-destructive and closed-world, and the description still adds substantive context on top: no data retention or logging, no cleaned file returned, pattern-based matching that does not verify mailbox existence, and hard input limits. It also discloses the exact normalization rules (case- and whitespace-insensitive email matching, placeholders counted as blank), which is behavior an agent could not infer from the 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?

The core constraint and detection-only qualifier are front-loaded, and the long finding enumeration is a single dense but scannable run. Every listed category corresponds to actual output content, so nothing is filler. It is on the long side and could be split into a bullet-style list for faster parsing.

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?

With no output schema, the description carries the burden of telling the agent what comes back, and it does so by enumerating the finding categories plus a mechanical phone fix suggestion. What is missing is the shape of the result (per-row versus aggregate counts, severity ordering) and guidance for CSVs exceeding the 500-row cap.

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%, so the baseline is 3, but the description adds real meaning: it defines what counts as blank for required columns (N/A, null, '-'), explains that phone fixes are only suggested 'when the fix is mechanical', and implies the role of default_country_code through the '+ country code and digits only' rule. It stops short of explaining column auto-detection or the delimiter handling, which remain schema-only.

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?

Specific verb plus resource: it checks a contact/lead CSV before CRM import and enumerates the finding categories it produces (duplicate emails, invalid syntax, role inboxes, blank required fields, phone format, geo drift, date formats, file problems). An agent knows immediately what the tool does. It never mentions the sibling preflight_sample, so it does not fully differentiate itself from that alternative.

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?

Placement in the workflow is explicit ('before a CRM import') and the boundary is drawn sharply: 'Detection only: it changes nothing, returns no cleaned file, and does not store or log your CSV.' The row/size limits (500 rows, 64 KiB) tell the agent when the input is too large. The one gap is that it never says to use preflight_sample instead when the input exceeds those limits.

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

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

Annotations already declare readOnly, idempotent, non-destructive and closed-world, so the safety profile is covered. The description still adds real value beyond them: the data is fake placeholder content, the response includes both the sample and its report, and nothing is stored or logged — a privacy guarantee the annotations do not express.

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?

A single front-loaded sentence that covers action, resource, sibling comparison, return shape, and privacy note with no filler. Every clause earns its place.

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?

With zero parameters and no output schema, an agent needs to know what it gets back and whether calling it has side effects — the description covers both (sample plus report, detection only, nothing stored). Nothing material is missing.

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?

There are no parameters, so per the baseline this scores 4; there is nothing for the description to disambiguate. The description correctly does not invent input requirements.

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 action (runs the same checks as preflight_csv) on a specific resource (the checker's built-in sample CSV) and explicitly contrasts it with the sibling that operates on user-supplied files. An agent can distinguish this from preflight_csv without opening either 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?

"so you can see the report format without sending a file" gives a clear selecting condition versus preflight_csv, and the sibling is named directly. It stops short of an explicit "do not use this for real files" statement, but the implicit routing is unambiguous.

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