ralph-install

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches a project scaffold from the author's official GitHub repository (github.com/tolulawson/ralph-harness) as specified in SKILL.md.
  • [COMMAND_EXECUTION]: The workflow involves cloning or fetching remote repository content to the local system to initialize the harness.
  • [PROMPT_INJECTION]: The skill establishes an indirect prompt injection surface by designating remote content as authoritative.
  • Ingestion points: Remote content is fetched from github.com/tolulawson/ralph-harness (SKILL.md).
  • Boundary markers: None; the agent is instructed to treat INSTALLATION.md as the source of truth.
  • Capability inventory: Performs repository cloning and file system operations.
  • Sanitization: None; the agent processes remote instructions directly to complete the setup.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 10:52 PM
Security Audit — agent-trust-hub — ralph-install