prune-prompt-pollution
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of markdown instructions aimed at improving the clarity and effectiveness of agent-facing documentation. It does not include any scripts, tool definitions, or command executions.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and refactor external documents, which represents an indirect prompt injection surface. However, the skill lacks high-risk capabilities like network access or file-system modification.
- Ingestion points: The skill processes rules files (AGENTS.md, CLAUDE.md), skill descriptions, prompts, and design documents as described in SKILL.md.
- Boundary markers: No specific delimiters are defined to separate the instructions from the data being processed.
- Capability inventory: There are no tool calls, network operations, or subprocess executions defined in the skill.
- Sanitization: No sanitization or validation of the input text is prescribed beyond refactoring its logical structure.
Audit Metadata