pipa-requirements-brief

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes content from untrusted external environments (e.g., chat logs, tickets, and cloud documents) which constitutes a surface for indirect prompt injection attacks.
  • Ingestion points: Data is pulled from Slack, Notion, Drive, and Linear via tool calls (SKILL.md, evals.json).
  • Boundary markers: The instructions mandate the separation of source facts from inferred requirements and the use of explicit status labels (e.g., used, partial, stale) to maintain data integrity.
  • Capability inventory: The skill possesses read/write capabilities for external workspace tools managed through composio-mcp.
  • Sanitization: A critical mitigation is implemented through a mandatory human-in-the-loop (HITL) gate, requiring explicit user approval for any external write operation after showing the scoped changes.
  • [DATA_EXPOSURE]: The skill reads local configuration files stored in the user's home directory (~/.pipa/communication-style.md and ~/.pipa/CONNECTORS.md). This is a standard pattern for managing application-specific settings and connector preferences and does not represent a security risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 08:06 AM
Security Audit — agent-trust-hub — pipa-requirements-brief