@oked/claude-code@0.6.22
OKed for Claude Code - zero-code human approval for sensitive actions
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe flagged agent-configuration write is performed by the explicit `oked init` command. The reviewed credential use sends the OKed API key and approval context to the configured OKed backend for the package’s stated approval function.
Decision evidence
public snapshot- The CLI changes Claude Code settings only in its explicit `init` command; package installation has no preinstall, install, or postinstall hook.
- The SDK uses its configured OKed API key and backend to submit an approval request to `/api/v1/approve`, matching the package’s stated approval purpose.
- The SDK’s default backend is `https://api.oked.ai`, and its configuration reads OKed-specific key settings from `OKED_API_KEY` or `~/.oked/config.json`.
- `init` writes an `oked-hook` entry to the project’s `.claude/settings.json` and can pair the device; users must explicitly run the command.
- When the installed hook handles an approval, the SDK sends approval context to the configured OKed backend. This is the package’s core function.
Source & flagged code
5 flaggedRuntime or CLI source writes behavior-bearing configuration into a user or project AI-agent control surface.
dist/cli.jsView on unpkgThe CLI changes Claude Code settings only in its explicit `init` command; package installation has no preinstall, install, or postinstall hook.
dist/cli.jsView on unpkg · L219Findings
Affected versions and remediation
This report applies to @oked/claude-code@0.6.22.
See all versions of @oked/claude-code for other recorded verdicts.
- Review the evidence and your use of @oked/claude-code@0.6.22 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.