registry / @netnodeag/kraftwerk / 0.72.0
@netnodeag/kraftwerk@0.72.0
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 17d ago · by lpm-firewall-aiThe inspected source supports normal agent-workspace functionality: its publication hook builds the package, and reviewed file, process, and network operations belong to workspace features. The Unicode finding is in editor text-handling code, and no concrete malicious install-time behavior was identified.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
The manifest hook runs before publication; the reviewed persistence and process actions are exposed through product operations.
Impact
No harmful effect from an unconsented install-time or import-time action was established.
Mechanism
Reviewed code handles editor text, workspace events, local runs, configured preview commands, remote repositories, and explicit daemon setup.
AI rationale
The inspected source supports normal agent-workspace functionality: its publication hook builds the package, and reviewed file, process, and network operations belong to workspace features. The Unicode finding is in editor text-handling code, and no concrete malicious install-time behavior was identified.
Decision evidence
public snapshotAI called this Clean at 93.0% confidence as Benign with high false-positive risk.
Evidence for AI clean decision
- package.json defines a prepublishOnly build hook; it does not define an install-time hook.
- Agent skills are listed and saved as named files through the agent-skill API, as product functionality.
- The client opens event streams for chat and vibeable topics, and run listing reads local output directories.
- The flagged editor bundle contains Unicode characters in text-direction and character-handling code; its sibling bundles include the editor and parser assets.
- Remote repository resolution, configured preview commands, workflow runs, and daemon installation are exposed through product operations.
Evidence against
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNetworkShellWebSocket
HighEntropyStringsMinifiedPossibleObfuscation
Source & flagged code
8 flaggeddist/remote.jsView file
1import { spawnSync } from "node:child_process";
L2: import { mkdir, stat } from "node:fs/promises";
High
dist/core/vibeables.jsView file
352cwd: r.dir,
L353: shell: true,
L354: detached: process.platform !== "win32",
High
bin/kraftwerk.jsView file
15if (existsSync(src) && process.env.KRAFTWERK_DIST !== "1") {
L16: const { register } = await import("tsx/esm/api");
L17: register();
Findings
1 Critical4 High5 Medium7 Low
CriticalTrojan Source Unicodeweb/dist/assets/editor-6D-WNjZY.js
HighChild Processdist/remote.js
HighShelldist/core/vibeables.js
HighRuntime Package Installdist/core/runner.js
HighSemantic Analysis Limiteddist/cli/daemon-login.js
MediumDynamic Requirebin/kraftwerk.js
MediumNetwork
MediumEnvironment Vars
MediumInstall Persistencedist/cli/daemon-login.js
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowEval
LowWeak Cryptodist/core/instances.js
Affected versions and remediation
This report applies to @netnodeag/kraftwerk@0.72.0.
See all versions of @netnodeag/kraftwerk for other recorded verdicts.
- Review the evidence and your use of @netnodeag/kraftwerk@0.72.0 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.