AI Security Review
scanned 2h ago · by lpm-firewall-aiReview flagged AI-agent configuration or capability changes. This remains warn-only unless evidence shows foreign-agent hijack through preinstall/install/postinstall, hidden persistence, exfiltration, remote code execution, or other concrete malicious behavior.
Decision evidence
public snapshot- `bin/run.mjs` default `npx youmindag` path overwrites project `AGENTS.md` and copies `.opencode/` configuration, plugins, and skills.
- `template/.opencode/opencode.json` activates `plugins/context-loader.js` in the target project.
- `context-loader.js` runs later in OpenCode, logs local task/tool metadata, rewrites bash commands with `echo` prefixes, and conditionally runs local `npx graphify` and checkpoint scripts.
- `bin/run.mjs` installs `@sentropic/graphify` and invokes Graphify through npm/npx after explicit CLI use.
- `package.json` has no `preinstall`, `install`, `postinstall`, or other lifecycle hook.
- No source network client or hard-coded exfiltration endpoint was found; URLs are metadata/schema references.
- File access targets the current project and local `.youmindag` state, not home-directory credentials or broad system paths.
- No eval, remote payload download, stealth persistence, or destructive cleanup chain was found.
Source & flagged code
3 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
bin/run.mjsView on unpkgPackage source invokes a package manager install command at runtime.
bin/run.mjsView on unpkg · L2