openchamber-sogawugu@1.21.3
[![GitHub stars](https://img.shields.io/github/stars/openchamber/openchamber?style=flat&logo=data%3Aimage%2Fsvg%2Bxml%3Bbase64%2CPHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIzMiIgaGVpZ2h0PSIzMiIgZmlsbD0iI2YxZWNlYyIgdmlld0JveD0iMCAwIDI1Ni
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package has no automatic lifecycle hook, but it exposes runtime creation of AI-agent instruction files and contains suspicious AI-directed text in a bundled asset. This is an unresolved agent-control-surface risk rather than a proven automatic malware chain.
Decision evidence
public snapshot- The bundled frontend contains direct instructions to a “working agent”, which is reviewer/AI-directed content.
- Runtime skill creation can write behavior-bearing skill files to user or project agent directories.
- A POST route invokes skill creation from request-supplied configuration.
- The manifest exposes server and CLI entrypoints but has no automatic install lifecycle hook.
- No preinstall, install, or postinstall script is declared.
- Startup persistence is reached only through the explicit `openchamber startup enable` command.
- The observed Cloudflare request is a connectivity check for a user-invoked tunnel feature.
Source & flagged code
12 flaggedFindings
Affected versions and remediation
This report applies to openchamber-sogawugu@1.21.3.
See all versions of openchamber-sogawugu for other recorded verdicts.
- Review the evidence and your use of openchamber-sogawugu@1.21.3 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.