Official Anchore Grype MCP Server for AI-assisted vulnerability scanning
This server has two tools with extremely minimal definitions. Both tools lack input schemas (empty {} objects), descriptions are generic and under-informative, and there is no structured output documentation. The naming is reasonable but the overall schema and parameter quality is critically deficient. Per the hard rules: tools with NO input schema must score 0 on schema; descriptions under 50 chars should score very low. This server falls well below production baseline (194 chars average, 100% of A+ tools documented).
Find the grype binary and check if it's installed. Returns information about grype installation status.
Install or update the grype binary to the latest version. Downloads and installs grype using the official Anchore installer.
Both tools have empty input schemas ({}). No parameters are defined, typed, or constrained.
Tool descriptions are extremely brief (29 and 36 chars respectively). Baseline average is 194 chars. Descriptions lack context on WHEN to call, WHAT is returned, and WHY the agent should choose this tool over alternatives.
No output schema is documented. Callers cannot know what structure find_grype returns (path string? JSON object with status?). update_grype returns boolean? Object?
update_grype performs a WRITE operation (downloads and installs software) but the description does not explicitly state this is destructive or irreversible. LLMs need to know which calls are safe to retry.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 37 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 23 | - | v1 |
No error handling guidance. If grype is not installed, find_grype will fail, what should the agent do next? Should it call update_grype?
update_grype downloads from a remote URL and executes shell scripts. No mention of validation, timeouts, or fallback behavior.