pipa-project-charter

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXFILTRATION]: The skill accesses application-specific configuration files such as ~/.pipa/CONNECTORS.md. This access is restricted to the skill's own configuration environment and does not involve accessing sensitive system credentials or user files.
  • [COMMAND_EXECUTION]: External tool interactions via composio-mcp are governed by strict verification steps, requiring the agent to validate tool schemas and confirm active permissions before processing any external data.
  • [PROMPT_INJECTION]: The skill incorporates defensive instructions that prioritize core safety and approval logic over formatting or presentation files, mitigating potential configuration-based overrides.
  • [DATA_EXFILTRATION]: A mandatory human-in-the-loop approval gate is enforced for all external write operations, requiring the agent to present exact changes to the user before execution to prevent unauthorized data modification.
  • [PROMPT_INJECTION]: The skill ingests data from external SaaS applications (Notion, Jira, Linear). While this creates a potential indirect prompt injection surface, the risk is mitigated by explicit approval gates for all state-changing operations and a requirement to verify tool schemas before processing data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 08:06 AM
Security Audit — agent-trust-hub — pipa-project-charter