Skills
skills/dedalus-erp-pas/foundation-skills/GitLab Code Review/Gen Agent Trust Hub

GitLab Code Review

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data from GitLab merge requests, creating an attack surface for indirect prompt injection.\n
  • Ingestion points: Data is fetched via gitlab-mcp tools such as get_merge_request_diffs and get_file_contents as specified in SKILL.md.\n
  • Boundary markers: Absent; instructions do not provide delimiters or directives to ignore instructions within the retrieved content.\n
  • Capability inventory: The skill has the ability to write to the platform via create_note and create_merge_request_thread tools in SKILL.md.\n
  • Sanitization: Absent; no evidence of filtering or validation of the ingested content exists in the skill definition.\n- [DATA_EXFILTRATION]: The skill performs network operations to a specific self-hosted GitLab instance.\n
  • Evidence: Interacts with https://gitlab-erp-pas.dedalus.lan using gitlab-mcp tools in SKILL.md. This domain is consistent with the author's organizational identity.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 10:37 PM