skills/iuliandita/skills/roadmap/Gen Agent Trust Hub

roadmap

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external Git repositories during its competitive scan workflow.
  • Ingestion points: The skill ingests issue titles, PR descriptions, and community comments from third-party repositories using gh or glab commands as described in references/competitive-scan.md.
  • Boundary markers: The instructions lack specific technical delimiters (e.g., XML tags or clear-text boundaries) to isolate external content, which could allow maliciously crafted issue titles or comments to influence the agent's behavior.
  • Capability inventory: The skill performs shell command execution (git, gh, glab) and has local file write access to the ROADMAP.md file.
  • Sanitization: While the suggested GitHub Action in references/trigger-integration.md correctly uses environment variables to mitigate script injection, the primary agent logic for parsing and matching external features does not include explicit sanitization or filtering of control characters from the ingested data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 08:18 AM
Security Audit — agent-trust-hub — roadmap