Skip to main content
Glama

Fairquery Agent Trust & Data Safety

by com.getfairquery
OverviewDiscussionsAdmin

Check package versions for known vulnerabilities (free, OSV.dev)

osv_query_packages
Read-onlyIdempotent

WHAT: Query the public OSV.dev database (api.osv.dev querybatch) for exact package versions; returns which are vulnerable and the advisory IDs. WHEN: Quick "is lodash 4.17.20 vulnerable?" checks for up to 200 packages; then osv_get_vulnerability for details of specific IDs. WHEN NOT: Parsing manifests/lockfiles, severity filtering, fixed/recommended versions in one shot -> osv_audit_dependencies (paid). COST: Free; no token. Package names and versions leave the client, are received by this Worker, and are forwarded to api.osv.dev. Nothing is sent to Apify. This Worker's code is designed not to store inputs or results. RETURNS: Paged "package" rows (vulnerable, vuln_ids) or per-entry INVALID_ECOSYSTEM "error" rows. Tool errors: OSV_FETCH_FAILED (retryable). BEHAVIOR: Runs on this Fairquery-hosted Worker (Cloudflare), not on the caller's machine; inputs leave the client. This Worker's code is designed not to store inputs, results, or an Apify token. One HTTPS request to api.osv.dev (Google-run public API) for all packages; typically <2 s, 20 s timeout. Results change only when OSV publishes advisories. PARAMETERS: packages: each entry is checked independently and returns exactly one row, in input order. An unknown ecosystem yields a per-entry INVALID_ECOSYSTEM error row while the other entries still run. A version range such as "^4.17.0" or ">=2" is not resolved and normally returns no vulnerabilities (a false "safe"), so pin versions from your lockfile first. Duplicate entries are queried and returned twice. limit/offset page the rows (all packages are sent to OSV in a single request); meta.vulnerable_count counts all packages, not only the page. EXAMPLE: {"packages":[{"ecosystem":"npm","name":"lodash","version":"4.17.20"}]} -> vulnerable true, vuln_ids ["GHSA-35jh-r3h4-6jhm",...].

Input Schema

NameRequiredDescriptionDefault
limitNoOptional page size: items to return, integer 1-200 (default 100). Example: 50
offsetNoOptional zero-based number of items to skip, integer 0-1000000 (default 0); pass nextOffset from the previous response, e.g. 100
packagesYesRequired. 1-200 {ecosystem, name, version} objects, one exact version each. Example: [{"ecosystem":"npm","name":"lodash","version":"4.17.20"},{"ecosystem":"PyPI","name":"django","version":"3.2.0"}]

Output Schema

NameRequiredDescriptionDefault
metaNoTool-specific context (summary counts, cost notes, ...)
countYesNumber of items in this page
itemsYesResult rows for this page
totalYesTotal items before paging
offsetYesOffset of the first item in this page
hasMoreYesTrue if more items exist after this page
nextOffsetYesOffset to pass for the next page, or null

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.9/5.0
Behavior5/5

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

Annotations already declare readOnly/idempotent/openWorld, and the description goes well beyond them: data-flow disclosure, single HTTPS request to api.osv.dev, <2s typical / 20s timeout, retryable OSV_FETCH_FAILED error code, and the fact that results change only when OSV publishes advisories. That is exactly the extra context the lower bar for annotated tools asks for.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The WHAT/WHEN/WHEN NOT/COST/RETURNS/BEHAVIOR/PARAMETERS/EXAMPLE labeling makes it skimmable and front-loaded. It loses a point for redundancy between COST and BEHAVIOR, which both restate that inputs leave the client and that the Worker does not store inputs or results.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With an output schema present, return-format detail is optional, yet the description still summarizes row shapes and error rows. Latency, failure modes, paging caveats, and the false-safe pitfall on version ranges are all covered, leaving nothing an agent needs to call this correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline would be 3, but the description adds semantics the schema cannot express: each entry is checked independently and returns exactly one row in input order, unknown ecosystems yield per-entry INVALID_ECOSYSTEM rows without failing the batch, ranges like '^4.17.0' silently return a false 'safe', duplicates are returned twice, and meta.vulnerable_count is not page-scoped.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The WHAT line states a specific verb and resource (query the OSV.dev database for exact package versions) plus the concrete return payload (vulnerable flag and advisory IDs). It explicitly names the two sibling tools it is not: osv_get_vulnerability for details and osv_audit_dependencies for manifest parsing.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Dedicated WHEN and WHEN NOT sections give the positive trigger (quick 'is lodash 4.17.20 vulnerable?' for up to 200 packages), the explicit exclusions (manifest/lockfile parsing, severity filtering, fixed versions) and route those cases to osv_audit_dependencies. Nothing is left to inference.

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

All Fairquery Agent Trust & Data Safety toolsInspect
Try in Browser
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources