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

With readOnlyHint and destructiveHint=false already covering safety, the description adds substantial behavior an agent cannot get from annotations: payloads carry no description text, first EPSS score emits epss_changed, later changes only on band crossings with 0.8 downward hysteresis, and applicability is a flattened NVD configuration list where vulnerable=false is context, not exclusion. These are non-obvious semantics that materially affect interpretation of results.

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?

Nearly every clause carries real technical content, but the presentation is a dense semicolon-chained paragraph that buries the tool's core purpose behind caveats and EPSS event details. Front-loading the what-it-returns statement before the applicability/hysteresis minutiae would substantially improve scannability without cutting content.

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 two-parameter, no-output-schema tool, the description covers the essential operational context: pagination continuation caveats, watermark replay reconciliation, rejection/withdrawal semantics, and expiry recovery. The main gap is that the return payload shape (fields available per record beyond applicability) is only partly implied, which an agent would need to discover empirically.

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 100%, so both limit and page_token are already documented in the schema, setting the baseline at 3. The description adds some value by noting that 'Later snapshot pages may contain revisions newer than the watermark,' which contextualizes page_token ordering, but says nothing about limit behavior or defaults, so no lift above baseline.

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 never crisply states what the tool returns in a single verb+resource sentence; the closest it gets is 'Public vulnerability records only' plus indirect references to 'snapshot records' and 'Later snapshot pages.' An agent can infer it is a paginated snapshot of public vulnerability records, but the purpose is embedded in caveats and workflow instructions rather than front-loaded, and differentiation from siblings like search_cve or get_vulnerability_changes is only partial.

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?

Strong operational guidance: 'Load the snapshot, replay get_vulnerability_changes with source=all from watermark_cursor, keep the higher revision per cve_id' explicitly names the companion tool and the sequence. It also states recovery behavior ('On cursor_expired or snapshot_expired restart the snapshot') and routes daily EPSS values to get_epss_score, but it does not say when NOT to use this tool versus search_cve or other query siblings.

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