skills/parcha-ai/parcha-skills/recap/Gen Agent Trust Hub

recap

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill follows security best practices by enforcing owner-only file and directory permissions (0600/0700) for all generated session manifests and event ledgers.
  • [SAFE]: A dedicated privacy module (scripts/privacy.py) provides robust, deterministic redaction of API keys, tokens, and private keys across a wide range of providers, ensuring sensitive data is not leaked into session summaries.
  • [SAFE]: External command execution via git and dependency scripts is performed safely using non-shell subprocess calls with hardcoded, read-only arguments, effectively preventing command injection vulnerabilities.
  • [SAFE]: The skill incorporates clear instructional guardrails (a "truth contract") that prevent the agent from treating ingested transcript data as executable instructions, mitigating the risk of indirect prompt injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 03:13 AM
Security Audit — agent-trust-hub — recap