@passioncode-ai/fabric-agent-adapter@0.47.1
Create Fabric-compatible agents, run them as always-alive local services with dashboards (fabric-service/0.1), and adapt existing projects to the Fabric Agent Contract. This package is the installer CLI.
AI Security Review
scanned 13d ago · by lpm-firewall-aiThe manifest has no npm lifecycle hook, and the flagged file operations are tied to explicit CLI use or the package's local service and session features. The session observer is gated on the registered Fabric service and sends limited facts rather than file contents.
Decision evidence
public snapshot- package.json has a test script and CLI entry point, with no npm install lifecycle hook.
- The launcher reads Claude plugin metadata and copies bundled skills into user skill directories only when the CLI runs.
- Claude session hooks send limited session and Git facts only when Fabric is registered, running, and advertises session-hooks/1; the observed facts omit file paths and contents.
- The session observer sends those facts through the registered Fabric MCP connection. The flagged supervisor code manages local service units.
- The identifier helper normalizes strings, and the protocol helper checks event data for credential-shaped values.
Source & flagged code
5 flaggedManifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.
bin/fabric-agent-adapter.jsView on unpkgPackage source references dynamic require/import behavior.
bin/fabric-agent-adapter.jsView on unpkg · L14Findings
Affected versions and remediation
This report applies to @passioncode-ai/fabric-agent-adapter@0.47.1.
See all versions of @passioncode-ai/fabric-agent-adapter for other recorded verdicts.
- Review the evidence and your use of @passioncode-ai/fabric-agent-adapter@0.47.1 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.