Known vulnerabilities for exact package versions from OSV, with fixes. Paid per call, x402.
Scored across 1 tool
Only one tool exists, so there is zero ambiguity. Agents cannot misselect because there is only one action available.
The single tool 'check_packages' follows a clear verb_noun pattern that directly describes its function. No naming inconsistencies are possible.
One tool is slightly on the low end, but the server is purpose-built for a single, comprehensive task (vulnerability lookup). The tool is substantial and not trivial, so the count feels appropriate for the scope.
The tool covers a wide range of ecosystems and provides detailed advisory information including severity, fixes, and dates. It fully fulfills its stated purpose of checking exact package versions against OSV data, with no obvious gaps for that narrow domain.
Look up the known vulnerabilities affecting exact package versions in the public OSV database (npm, PyPI, Go, Maven, crates.io, RubyGems, NuGet, Packagist, Pub, Hex), 1 to 50 packages per call. Per package: every OSV advisory for that version with its id, CVE/GHSA aliases, summary, severity (database label, CVSS vectors, v3 base scores), the versions that fix it or "no fix published", published/modified dates and a link. Plus a summary: packages checked, packages vulnerable, total advisories and the highest severity. Evidence, not advice: absence from OSV does not prove safety.
| Name | Required | Description | Default |
|---|---|---|---|
| packages | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden, and it does so thoroughly. It discloses the full output shape: advisory IDs, aliases, summaries, severity details, fix versions or 'no fix published', dates, and a summary. It also explicitly cautions that absence from OSV is not proof of safety, which is a meaningful behavioral limitation for an AI agent to understand.
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 a single dense paragraph that front-loads the primary action and scope, then systematically lists input constraints, output contents, and a caveat. Every sentence contributes either to call construction or expectation setting, with no filler or repeated schema 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?
Given the tool has no output schema unreplaced, no annotations, and one nontrivial nested parameter, the description is unusually complete. It explains exactly what will be returned per packageeb, what the aggregate summary includes, what the input limits are, and how to interpret the results. An agent has everything needed to invoke this tool correctly and understand its limitations.
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 description adds important semantics beyond the input schema by bounding calls to 1-50 packages and emphasizing exact versions rather than ranges. It also enumerates the supported ecosystems, which helps an agent construct valid requests. The schema already documents per-field details like name format and the 'exact version' requirement, so the description supplements rather than replaces that information.
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 specific verb and resource: 'Look up the known vulnerabilities affecting exact package versions in the public OSV database.' It names the supported ecosystems and clearly distinguishes that this is an evidence-gathering tool, not an advice-giving one. This leaves no ambiguity about what the tool does.
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?
It clearly conveys usage constraints: callers must specify exact versions, can check 1 to 50 packages per call, and should treat results as evidence only. It also states a key limitation—absence from OSV does not prove safety. It does not name alternative tools or explicit when-not-to-use conditions, but no siblings are provided and the context is otherwise clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
check_packagesLook up CVEs, CISA KEV, EPSS and open-source package vulnerabilities, including malicious packages.
OSV.dev — Google's open-source vulnerability database
CVE, KEV, EPSS, SBOM and advisory lookups; per-CVE exposure from Shodan data (© Shodan). Keyless.
CVE lookups (NVD) and dependency-manifest audits (OSV) for AI agents. No API keys.
Add one secure layer between your agents and this server.