sred-project-organizer

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes third-party data from GitHub pull requests, Linear tickets, and Notion documents which can contain user-generated content or adversarial text.
  • Ingestion Points: Ingestion occurs in SKILL.md during Step 1, Step 2, and Step 5, where the agent reads data from user-provided Notion Work Summaries, related project PRs, and task tickets.
  • Boundary Markers: No explicit boundary delimiters or strict instruction isolation techniques are specified to separate the trusted system prompt from untrusted data inputs.
  • Capability Inventory: The skill interacts with the user's environment using Notion and Linear Model Context Protocol (MCP) integrations alongside the GitHub CLI tool to read issues/PRs and create new private documents.
  • Sanitization: Input validation and text sanitization mechanisms are absent, meaning embedded instructions in source code comments, PR descriptions, or ticket notes would be processed directly by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 08:28 AM
Security Audit — agent-trust-hub — sred-project-organizer