@theronap/agnoclast-mcp@0.9.290
Connect your AI assistant to Cortex — your org's projects, activity, gaps, and directives, scoped to you.
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe manifest has no install lifecycle hooks, and the flagged persistence and agent-configuration mutations are invoked through dedicated user commands. The credential-to-foreign-host hint is a false positive: the OpenAI URL is used to read a JWT claim locally, while the proof is sent to the configured Agnoclast service as part of the confidential-brain feature.
Decision evidence
public snapshot- The manifest has no install, preinstall, or postinstall hook; its scripts are release, promote, and rollback commands.
- Background closeout is described as off by default and is installed by the explicit closeout-install command.
- The closeout job reviews quiet sessions with limits and runs using the user's Claude account.
- Channel configuration edits are part of the explicit channel command; with no argument, it reports status.
- The OpenAI URL appears as a JWT claim namespace parsed locally; the proof is sent to the configured Agnoclast service for a confidential-brain seat.
Source & flagged code
16 flaggedSource file is highly similar to a previously finalized malicious package; route for source-aware review.
lib/capture.mjsView on unpkgPackage source references weak cryptographic algorithms.
lib/capture.mjsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @theronap/agnoclast-mcp@0.9.290.
See all versions of @theronap/agnoclast-mcp for other recorded verdicts.
- Review the evidence and your use of @theronap/agnoclast-mcp@0.9.290 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.