skills/simpleeve/eo-skills/eo-archive/Gen Agent Trust Hub

eo-archive

Pass

Audited by Gen Agent Trust Hub on Nov 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes internal project documentation to verify archival criteria, which presents a surface for potential indirect prompt injection.
  • Ingestion points: The agent reads verification status, evidence, and manual acceptance items from local files such as change.md, evidence.md, and acceptance.md.
  • Boundary markers: The skill relies on specific markdown section headers (e.g., §2, §6) and YAML frontmatter fields for structure but does not implement explicit safety delimiters or instruction-ignore markers for the ingested content.
  • Capability inventory: The workflow involves executing git status, git commit, and git push, as well as calling vendor tools eo-sync and eo-doc-manager.
  • Sanitization: The risk is mitigated by mandatory human-in-the-loop confirmation steps, such as explicit confirmation prompts and delivery reports, before any significant state changes or commits occur.
  • [COMMAND_EXECUTION]: The skill utilizes standard version control commands (git status, commit, push) and local project management utilities (eo-sync, eo-doc-manager) to perform its archival duties. These operations are context-appropriate and involve user oversight.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 14, 2026, 09:54 AM