opencode-handoff-p2p

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a defense-in-depth security model for handling untrusted data from peer-to-peer transfers.
  • [COMMAND_EXECUTION]: The skill uses shell commands (git, gh, python) which are necessary for its primary purpose of managing GitHub-based inboxes. It mitigates command injection risks by using strict regex validation for all inputs (filenames, URLs, usernames) and mandatory single-quoting of shell arguments.
  • [DATA_EXFILTRATION]: While the skill involves network operations via the GitHub CLI and git, these are scoped to specific, user-configured private repositories. The configuration and trust files are stored in a skill-specific directory to prevent hijacking from other projects.
  • [PROMPT_INJECTION]: The skill explicitly identifies that fetched content is third-party data and mandates the use of a trust-boundary preamble. This preamble instructs the agent to treat the content as data rather than instructions, mitigating indirect prompt injection risks.
  • [REMOTE_CODE_EXECUTION]: The skill executes a local Python script (verify_inbox.py) for deterministic verification. This script follows secure coding practices, such as using subprocess.run with a hardened environment, avoiding system-wide git configurations, and validating all data at the byte level before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 04:00 AM
Security Audit — agent-trust-hub — opencode-handoff-p2p