github-agentic-workflows

Warn

Audited by Socket on Mar 1, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

This SKILL.md is largely documentation and examples for authoring GitHub Agentic Workflows. It does not contain executable malware or explicit exfiltration code. However, it describes and legitimizes powerful capabilities (installing a CLI extension, configuring engine API keys/PATs, enabling bash and web-fetch tools, and workflows that can write to the repository) that — if misconfigured or combined with untrusted inputs — create notable supply-chain and credential-forwarding risks. Recommend treating workflows and gh-aw extension installs as high-value supply-chain components: audit the gh-aw extension, use minimal secrets/scopes, avoid granting bash/web-fetch unless necessary, require human approval gates for write actions, and sanitize any untrusted input before letting an agent take write/exec actions.

Confidence: 75%Severity: 75%
Audit Metadata
Analyzed At
Mar 1, 2026, 09:50 PM
Package URL
pkg:socket/skills-sh/hack23%2Fhomepage%2Fgithub-agentic-workflows%2F@463e4b2e9f4c19d599db371db31ad2ab0405b90f