@netnodeag/kraftwerk@0.46.2
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-aiNo malicious install hook, credential theft, exfiltration, or covert persistence was found. The package nevertheless provides explicit agent-control mutation and shell execution through its inspector, so it warrants a warning rather than a clean verdict.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
A user or a caller able to access the inspector sends the dev-start or agent-management API request.
Impact
A malicious project configuration or unauthorized inspector caller could run commands in the project context or change agent behavior.
Mechanism
Shell execution of project configuration and project agent-prompt mutation.
AI rationale
No malicious install hook, credential theft, exfiltration, or covert persistence was found. The package nevertheless provides explicit agent-control mutation and shell execution through its inspector, so it warrants a warning rather than a clean verdict.
Evidence
dist/inspector/server.jsdist/inspector/vibeables.jsdist/inspector/agents.jsvibeables/<slug>agents/<slug>/agent.ymlagents/<slug>/system.md
Decision evidence
public snapshotAI called this Suspicious at 87.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The inspector exposes a POST route that starts a project-defined development command.
- That command is run through a shell with the full inherited environment.
- The inspector API writes agent configuration and system-prompt files in the selected project.
- The package can clone or refresh a user-supplied remote Git repository into its cache.
- The manifest has no automatic install lifecycle hook; its only publish hook is prepublishOnly.
Evidence against
- The shell execution is reached through an explicit inspector API action and uses the selected vibeable directory.
- Agent files are constrained to validated slugs beneath the configured project agents directory.
- No source inspected showed credential harvesting, secret exfiltration, stealth persistence, or an install-time payload.
- The inspector defaults to loopback binding outside containers.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetwork
Source & flagged code
8 flaggeddist/remote.jsView file
1import { spawnSync } from "node:child_process";
L2: import { mkdir, stat } from "node:fs/promises";
High
dist/inspector/vibeables.jsView file
•matchType = previous_version_dangerous_delta
matchedPackage = @netnodeag/kraftwerk@0.2.0
matchedIdentity = npm:QG5ldG5vZGVhZy9rcmFmdHdlcms:0.2.0
similarity = 0.458
summary = stored previous version shares package body but lacks this dangerous source file
Critical
Previous Version Dangerous Delta
This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/inspector/vibeables.jsView on unpkg546cwd: r.dir,
L547: shell: true,
L548: detached: process.platform !== "win32",
High
Findings
2 Critical4 High4 Medium6 Low
CriticalTrojan Source Unicodeinspector/dist/assets/editor-DJbeG376.js
CriticalPrevious Version Dangerous Deltadist/inspector/vibeables.js
HighChild Processdist/remote.js
HighShelldist/inspector/vibeables.js
HighRuntime Package Installdist/inspector/runner.js
HighSemantic Analysis Limiteddist/inspector/agents.js
MediumDynamic Requirebin/kraftwerk.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowWeak Cryptodist/inspector/instances.js
Affected versions and remediation
This report applies to @netnodeag/kraftwerk@0.46.2.
See all versions of @netnodeag/kraftwerk for other recorded verdicts.
- Review the evidence and your use of @netnodeag/kraftwerk@0.46.2 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.