mega-goal-prompt
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill files consist solely of Markdown instructions and reference templates. There is no executable code, network access, or sensitive file system interaction.
- [NO_CODE]: The skill bundle does not contain any scripts (e.g., Python, JavaScript) or package definitions, relying entirely on the agent's natural language processing capabilities.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text to populate a mega-prompt template, which is a standard surface for indirect prompt injection.
- Ingestion points: The skill gathers user input across several branches defined in SKILL.md (Outcome, Context, Success Criteria, etc.).
- Boundary markers: The input is formatted into a fixed template provided in references/mega-prompt-template.md which uses structured sections.
- Capability inventory: The skill itself possesses no capabilities (tools, shell commands, or network access) and only outputs text.
- Sanitization: No sanitization is performed on user inputs before they are placed into the final template output.
Audit Metadata