@theronap/agnoclast-mcp@0.9.225
Connect your AI assistant to Cortex — your org's projects, activity, gaps, and directives, scoped to you.
AI Security Review
scanned 25d ago · by lpm-firewall-aiThe inspected source shows no install lifecycle hook or concrete attack chain. Agent configuration and Antigravity persistence are tied to explicit setup or install commands and are package-aligned integrations, so the scanner findings do not establish malicious behavior.
Decision evidence
public snapshot- package.json defines a CLI entry point and has no preinstall, install, or postinstall hook.
- Codex and Cursor adapters detect their respective configuration directories; Cursor is capture-only and does not inject prompt context.
- Setup wiring is reached through an explicit CLI subcommand, and it backs up and merges agent configuration.
- Managed skills target Claude Code and Codex; uninstall removes package-managed entries and associated files.
- Antigravity wire() creates a launchd watcher only when invoked; its generated script targets a local development checkout and skips when no token is wired.
- Transcript lookup reads Claude project files; inspected behavior shows no harvesting or exfiltration chain.
- The package can configure agent hooks, install managed skills, and create Antigravity launchd files when its setup or install commands are invoked.
- The CLI and library contain network calls, but inspected sources did not establish an unrelated credential recipient or concrete attack behavior.
Source & flagged code
6 flaggedPackage source references weak cryptographic algorithms.
lib/capture.mjsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @theronap/agnoclast-mcp@0.9.225.
See all versions of @theronap/agnoclast-mcp for other recorded verdicts.
- Review the evidence and your use of @theronap/agnoclast-mcp@0.9.225 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.