wildcrane-runner@0.6.12
WildCrane's Local Runner — connects this computer to WildCrane and runs its jobs on your local Claude Code, Codex, OpenCode or Cursor.
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe launcher and bundled runner perform the package’s stated job: connect to WildCrane after sign-in and run tasks through local agents. The flagged provider-key references configure those agents, and the prepare script is a build step rather than an install hook that sends data.
Decision evidence
public snapshot- The documented purpose is to connect the computer to WildCrane and run tasks through local coding agents.
- The launcher starts the runner; the README says first launch opens a browser for sign-in.
- The bundled code labels the Anthropic key as stored locally and only adds configured secrets to the selected provider’s environment.
- Cloud requests use the runner session’s bearer token; no automatic forwarding of third-party model keys to an unrelated host was identified.
- The prepare script runs a build command; no preinstall, install, or postinstall script is declared.
- The runner automatically checks for cloud-directed version updates and downloads versions from the configured npm registry.
- The bundled runner can execute local agents and their workflows, which is central to its documented purpose.
Source & flagged code
10 flaggedSource appears to send environment or credential material to an external endpoint.
dist/akb/dist/kanban.mjsView on unpkg · L58Package source references child process execution.
dist/akb/dist/kanban.mjsView on unpkg · L1168Findings
Affected versions and remediation
This report applies to wildcrane-runner@0.6.12.
See all versions of wildcrane-runner for other recorded verdicts.
- Review the evidence and your use of wildcrane-runner@0.6.12 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.