opencode-skills-collection@4.0.75
OpenCode CLI plugin that automatically downloads and keeps skills up to date.
AI Security Review
scanned 26d ago · by lpm-firewall-aiThis is a first-party OpenCode plugin whose startup path mutates the OpenCode skills directory and its own skill vault, which is agent-extension setup rather than an install-hook hijack. No credential send, remote payload, or shell execution was found in the plugin entrypoint.
Decision evidence
public snapshot- The default plugin export runs on OpenCode startup and targets ~/.config/opencode/skills.
- installSkillsToVault copies bundled skill trees into ~/.config/opencode/skill-libraries and recursively deletes vault entries absent from the index.
- generatePointers writes category pointer SKILL.md files into the active skills directory and removes pointer directories that are no longer expected.
- package.json has no preinstall, install, or postinstall hook; scripts are only build, test, and prepublishOnly.
- No child_process, eval, dynamic import, or network client appears in dist.
- Pointer cleanup is limited to directories whose names end with -category-pointer.
- Skill copy paths reject .. and separators before cpSync.
- Patches apply only when the user-owned skill-filter.jsonc supplies them; the default patch list is empty.
Source & flagged code
14 flagged · loading sourcePackage contains a high-severity secret pattern.
bundled-skills/developer-signup-flow/SKILL.mdView on unpkg · L165Stripe test secret key in bundled-skills/developer-signup-flow/SKILL.md
bundled-skills/developer-signup-flow/SKILL.mdView on unpkg · L165Findings
Affected versions and remediation
This report applies to opencode-skills-collection@4.0.75.
See all versions of opencode-skills-collection for other recorded verdicts.
- Review the evidence and your use of opencode-skills-collection@4.0.75 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.