agency-lang@0.22.0
The Agency language
AI Security Review
scanned 24d ago · by lpm-firewall-aiStatic hints pointed to process execution and network use, but inspected code tied these behaviors to package runtime functions and the user-facing CLI. The manifest scripts excerpt shows no install lifecycle hook, and no concrete attack behavior was established.
Decision evidence
public snapshot- package.json contains development scripts and no install lifecycle hook.
- The IPC helper forks a child process with configured memory limits and an IPC channel.
- OAuth code posts a token exchange to its supplied OAuth endpoint.
- The ffmpeg helper launches ffmpeg to process input data.
- The CLI launches compiled user output through Node with shell execution disabled.
Source & flagged code
9 flaggedPackage source references child process execution.
dist/lib/runtime/ipc.jsView on unpkg · L9Package source references shell execution.
dist/lib/stdlib/oauth.js#virtual:normalized:round1View on unpkg · L27Findings
Affected versions and remediation
This report applies to agency-lang@0.22.0.
See all versions of agency-lang for other recorded verdicts.
- Review the evidence and your use of agency-lang@0.22.0 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.