talk-stage5-script
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill functions as a structured workflow for content generation. It utilizes 'Read' and 'Write' tools to process local markdown files within a designated project directory ('talks/').
- [NO_CODE]: The skill consists solely of instructions and markdown templates. It does not include or execute any scripts (Python, JavaScript, shell, etc.), significantly reducing the attack surface.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external project files (summary, concepts, angles, etc.) to populate templates. While this represents a theoretical ingestion surface for indirect prompt injection, the risk is mitigated by the skill's requirement for manual user verification ('Stage 4 CHECKPOINT') and its output being intended for manual copy-pasting by the user rather than automated execution.
Audit Metadata