@netnodeag/kraftwerk@0.45.0
Open-source agentic workspace for teams — where people and AI agents get work done together: agents (Claude Code, Codex, Pi), skills, shared knowledge, repeatable workflows, and deterministic verification
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package is not shown to be malicious, but its explicit project-command execution and incompletely auditable minified inspector asset leave a real dual-use risk. There is no automatic lifecycle trigger or concrete malicious chain supporting a block.
Decision evidence
public snapshot- The package exposes a CLI and bundles a large minified inspector asset, which limits complete source-level auditing of that asset.
- Remote workflow resolution clones a supplied Git source and can hard-reset its cached checkout.
- The inspector launches `npx kraftwerk run` in the consumer project when a run is triggered.
- The inspector runs a project-configured development command through a shell with inherited environment variables.
- The agent editor writes agent configuration and system-prompt files in the configured project directory.
- package.json has only a prepublishOnly script, not an install-time lifecycle hook.
- The manifest does not runtime-depend on its own package name.
- The reviewed code contains no confirmed credential collection, outbound secret upload, or hidden persistence.
- Shell and process execution are tied to explicit inspector actions and project workflow configuration.
Source & flagged code
8 flaggedThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/inspector/vibeables.jsView on unpkgFindings
Affected versions and remediation
This report applies to @netnodeag/kraftwerk@0.45.0.
See all versions of @netnodeag/kraftwerk for other recorded verdicts.
- Review the evidence and your use of @netnodeag/kraftwerk@0.45.0 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.