ralph-upgrade

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill clones the Ralph harness scaffold from the author's repository at https://github.com/tolulawson/ralph-harness to acquire upgrade materials.\n- [COMMAND_EXECUTION]: The skill runs Python scripts (scripts/check-upgrade-surface.py and scripts/migrate-installed-runtime.py) from the newly cloned repository to manage the migration logic and project state.\n- [PROMPT_INJECTION]: The skill uses an external file, UPGRADING.md, as the authoritative guide for the agent's upgrade actions, creating a surface for indirect prompt injection.\n
  • Ingestion points: UPGRADING.md (fetched from GitHub).\n
  • Boundary markers: Absent; the file is treated as the primary source of truth for the workflow.\n
  • Capability inventory: Subprocess execution of Python scripts via the shell.\n
  • Sanitization: Absent; the agent is instructed to follow the external instructions directly.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 10:53 PM
Security Audit — agent-trust-hub — ralph-upgrade