Skills
Navigate
skills/ortus-boxlang/bx-sites-skills/bx-sites-actions/Gen Agent Trust Hub

bx-sites-actions

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub Action logs via the get_job_logs tool. This creates a potential surface for indirect prompt injection if log content contains malicious instructions intended to manipulate the agent. However, this access is inherent to the skill's purpose of troubleshooting CI failures.
  • Ingestion points: Workflow logs retrieved through get_job_logs in SKILL.md.
  • Boundary markers: No explicit delimiters are defined to separate log content from instructions.
  • Capability inventory: The skill can trigger workflow actions using actions_run_trigger (rerun, cancel).
  • Sanitization: The instructions do not specify sanitization for the log content before processing.
  • [COMMAND_EXECUTION]: The skill uses specialized tools (actions_list, actions_run_trigger) to interact with GitHub's API. These operations are limited to the specific repository ortus-boxlang/bx-sites and follow standard administrative workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 03:14 AM
Security Audit — agent-trust-hub — bx-sites-actions