AGENTSCORE-2026-0186
MCP package: open-memex
Published 11/7/2026 · 0.8.2 → 0.9.0
Automated, not yet reviewed
This advisory was generated automatically by a pattern-matching scanner when the package's score changed. Its findings have not been checked in the package source yet. Pattern matching produces false positives: when automated command-injection findings were first reviewed in source, most were not exploitable. Treat unreviewed findings as leads, not verdicts.
open-memex updated from 0.8.2 to 0.9.0. Score changed 70/100 to 50/100 (-20). Risk: MODERATE to ELEVATED. 5 findings.
70 → 50
Score
MODERATE → ELEVATED
Risk
WARN
Verdict
Findings
- low install_script: Package has 'postinstall' script: node scripts/postinstall.js
- low command_injection: Potential command injection: shell execution with template literal input (downgraded — mitigators detected in scope: sanitizer:.exec(`DROP, sanitizer:${SCHEMA_VERSION})
- critical hardcoded_secret: Hardcoded secret found (AWS key, OpenAI key, GitHub token, or npm token)
- low unsafe_eval: Uses eval() with dynamic input (downgraded — mitigators detected in scope: sanitizer:${m.id}, test_fixture:fixture)
- low no_provenance: Package is not published with provenance attestations or trusted publishing. Published by: stoneskin