gh-actions-validator
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and parses YAML files from the
.github/workflows/directory, which represents untrusted external data. While primarily used for security auditing, instructions or payloads embedded in these files (e.g., in comments or metadata) could theoretically influence the agent's logic. - Ingestion points: Reads workflow files in
.github/workflows/(referenced in SKILL.md and scripts/validate-workflow.sh). - Boundary markers: Absent; there are no delimiters or explicit instructions to ignore embedded commands within the analyzed files.
- Capability inventory: Bash (git, gcloud), Write, Edit, Grep, Glob, and Read access.
- Sanitization: The validation script uses grep on raw file content without sanitization or escaping mechanisms.
- [COMMAND_EXECUTION]: The skill utilizes the
gcloudCLI and shell scripts (setup-wif.sh,validate-workflow.sh) to perform IAM audits and configure Workload Identity Federation resources. These operations are aligned with the skill's primary purpose of security hardening.
Audit Metadata