@goobit/ai@1.0.0
Enterprise AI Git Collaboration Governance & Smart Skill Hub for Heterogeneous AI Agents
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe manifest activates an install hook that automatically writes behavior-bearing instructions into multiple consumer AI-agent configuration locations. This meets the policy for blocking unconsented install-time control-surface mutation.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
Installing the package in an external project with npm, which runs its postinstall hook.
Impact
Subsequent AI-agent sessions in the project can be steered by configuration written during dependency installation.
Mechanism
The hook calls setupGovernance, which writes AGENTS.md and redirects for Claude, Cursor, Gemini, and Copilot that instruct those tools to follow AGENTS.md and load bundled skills.
Policy narrative
Installing the package in an external project activates its postinstall hook, which calls setupGovernance without an affirmative setup prompt. The routine writes package-provided instructions into AGENTS.md and several AI-agent configuration files, directing those agents to follow the generated rules and load bundled skills. This is an unconsented install-time change to a broad AI-agent control surface.
AI rationale
The manifest activates an install hook that automatically writes behavior-bearing instructions into multiple consumer AI-agent configuration locations. This meets the policy for blocking unconsented install-time control-surface mutation.
Evidence
package.jsonbin/postinstall.jsbin/harness-bridge.jsAGENTS.mdCLAUDE.md.cursorrules.gemini/rules/project.md.github/copilot-instructions.md
Decision evidence
public snapshotAI identified this as a dangerous AI-agent capability at 97.0% confidence with low false-positive risk.
Evidence for policy risk
- The manifest runs bin/postinstall.js as an npm postinstall hook.
- The hook targets an external project from INIT_CWD and automatically calls setupGovernance.
- The setup writes generated governance instructions to the project root AGENTS.md.
- The generated redirect tells several AI tools to follow AGENTS.md and load package skills.
- The setup writes those redirects into Claude, Cursor, Gemini, and Copilot configuration files.
- The entrypoint describes bridging AI harnesses to AGENTS.md.
Evidence against
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystem
HighEntropyStrings
Source & flagged code
7 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node ./bin/postinstall.js
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
package.jsonView on unpkg•scripts.postinstall = node ./bin/postinstall.js
Medium
Ambiguous Install Lifecycle Script
Install-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgbin/harness-bridge.jsView file
Findings
1 Critical3 High4 Medium4 Low
CriticalAi Agent Control Hijackbin/harness-bridge.js
HighInstall Time Lifecycle Scriptspackage.json
HighTrigger Reachable External Ai Agent Control Surface Mutationbin/harness-bridge.js
HighSemantic Analysis Limitedpackage.json
MediumAmbiguous Install Lifecycle Scriptpackage.json
MediumEnvironment Vars
MediumShips Build Helperassets/skills/excel-reader/scripts/read_excel.py
MediumStructural Risk Force Deep Review
LowScripts Present
LowWeak Cryptobin/harness-bridge.js
LowFilesystem
LowHigh Entropy Strings
Affected versions and remediation
This report applies to @goobit/ai@1.0.0.
See all versions of @goobit/ai for other recorded verdicts.
- Avoid installing @goobit/ai@1.0.0. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.