@herjarsa/omo-meta-governor@0.31.8
Self-judging agent orchestration layer for OpenCode. Reads memory, scores sessions, dispatches decisions.
AI Security Review
scanned 12d ago · by lpm-firewall-aiThe source supports a warning for guarded agent-control behavior: the library can inject instructions to push commits upstream. No install lifecycle hook or direct package-side push was established, so the evidence does not support blocking publication.
Decision evidence
public snapshot- dist/lib.js can inject a synthetic user message instructing the agent to commit and push changes to local, fork, and upstream repositories.
- The prompt injection is gated by session state and a check for a plan; it can recur at most once.
- package.json describes an OpenCode agent orchestration package and defines no install lifecycle hook.
- The library has an auto-upgrade path for graph tooling, but it is separate from the agent prompt behavior.
- The package exposes a library and MCP executable; its manifest contains build, typecheck, and test scripts but no preinstall, install, or postinstall hook.
- The flagged agent integration locations inspect plan markers in AGENTS.md; the inspected code does not establish a write to that file.
Source & flagged code
7 flagged · loading sourcePackage source invokes a package manager install command at runtime.
dist/lib.jsFindings
Affected versions and remediation
This report applies to @herjarsa/omo-meta-governor@0.31.8.
See all versions of @herjarsa/omo-meta-governor for other recorded verdicts.
- Review the evidence and your use of @herjarsa/omo-meta-governor@0.31.8 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.