pkg-oracle — Dependency Trust Oracle
Server Details
Blocks typosquatted or hallucinated npm/PyPI packages before an AI agent installs them.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 1 tool
There is only one tool, so there is no possibility of misselection or overlapping purpose. Its scope (verify a package's existence, CVEs, scorecard, and typosquat risk) is clearly stated and self-contained.
verify_package follows a clean verb_noun snake_case convention. With a single tool there is no mixed convention to create inconsistency, and the name matches the described action exactly.
A single-tool surface is defensible for a narrowly scoped trust oracle that performs one composite check, but it sits below the typical 3-15 sweet spot and offers no alternate entry points. Nothing is redundant, so the count is lean rather than incoherent.
The core check (existence, OSV CVEs, OpenSSF Scorecard, typosquat detection) is thorough for one package, but obvious adjacent operations are missing: batch/manifest-wide verification, re-verification or cache status within a session, and any way to inspect the trust database the tool consults. Agents verifying a full dependency file must loop one call per package, which is a notable gap.
Available Tools
1 toolverify_packageVerify PackageAInspect
Dependency Trust Oracle. Call this BEFORE writing any package into a manifest (package.json, requirements.txt, pyproject.toml, Cargo.toml, ...). It checks whether the package actually exists on its registry, cross-references OSV.dev for known CVEs, pulls the package's OpenSSF Scorecard via deps.dev, and runs a Levenshtein-distance typosquat/slopsquat check against a curated list of popular packages combined with the package's publish age. Returns a synthetic verdict: ALLOW (no issues found), WARN (proceed with caution — read the findings before installing), or BLOCK (do not install — likely a hallucinated package name, an active typosquat, or a known critical/high-severity vulnerability). Always call this before running an install command for a package you have not already verified in this session. First 5 calls per caller are free; after that this tool requires x402 payment (USDC on Base) and will return a payment-required error with the amount and address to pay.
Inputs: name (required), ecosystem (required), version.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Verify npm packages before your AI agent installs them: hallucinations, advisories, API drift.
Check an npm/PyPI package before an AI agent installs it: 5 focused supply-chain tools.
Check packages for CVEs, slopsquatting, and CISA KEV before your AI agent installs them.
Package verdicts for AI coding agents: check npm, PyPI, Go and NuGet installs before they run.
Related MCP Servers
- AlicenseAqualityDmaintenanceVerifies npm package names for safety, detecting typosquats and nonexistent packages before your coding agent installs them.23 npmMIT
- AlicenseNot gradedqualityDmaintenanceValidates and checks packages across 19 ecosystems to prevent AI agents from installing hallucinated, deprecated, or malicious packages.60 npmAGPL 3.0
- AlicenseBqualityBmaintenancePre-install gate for AI coding agents. Checks npm, PyPI and crates.io package names before an install runs and blocks names that do not exist, were registered after models invented them, or sit one keystroke from a popular package. CLI, GitHub Action and MCP server. Apache 2.0.1306 PyPI1Apache 2.0
- AlicenseNot gradedqualityDmaintenanceEnables AI coding agents to automatically verify npm packages against the live registry before installation, flagging hallucinated, slopsquatted, or otherwise suspicious packages with risk verdicts.4 npmMIT
Glama MCP Gateway
Add one secure layer between your agents and this server.