skill-preflight
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes content from existing SKILL.md and agent configuration files (.claude/agents/*.md) using Read and Grep tools. This represents an indirect prompt injection surface where malicious instructions embedded in existing files could attempt to influence the 'Creation Guard' logic to incorrectly allow, block, or modify automation artifacts.
- Ingestion points: The skill instructions mandate reading files matching
skills/*/SKILL.mdand.claude/agents/*.md(documented in Step 2: Search Existing Artifacts). - Boundary markers: None identified. The skill does not instruct the agent to use delimiters or specifically ignore instructions found within the scanned artifacts.
- Capability inventory: The results of this analysis directly guide the agent's decision-making (PROCEED, EXTEND, COMPOSE, ITERATE, BLOCK) for new code creation and modification.
- Sanitization: No specific sanitization, filtering, or validation of the ingested content is performed before the analysis step.
Audit Metadata