prepare-walkthrough-presentation
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates strictly on local file transformation. All identified activities involve reading and writing markdown documents within a user-defined walkthrough directory.
- [DATA_EXPOSURE_AND_EXFILTRATION]: Analysis confirmed no network communication or access to sensitive system paths (e.g., credentials, SSH keys). It only reads from the specified walkthrough directory and writes to a presentation sub-folder.
- [PROMPT_INJECTION]: The instructions and metadata do not contain patterns intended to bypass safety filters or override agent behavior.
- [REMOTE_CODE_EXECUTION]: No commands for downloading or executing remote code were found. The workflow uses text-processing agents with no access to shell execution or dynamic evaluation tools.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data from the walkthrough artifacts. Ingestion points: Files like unified-walkthrough.md, sections/walkthrough-section-.md, and validation/validation-report-.md are read during Phase 1-4. Boundary markers: Absent; the instructions do not use specific delimiters or 'ignore embedded instructions' warnings for untrusted content. Capability inventory: Limited to file read/write operations for markdown content; no network operations, subprocess calls, or code execution (eval/exec) were detected in SKILL.md or agent-instructions.md. Sanitization: Absent; the skill relies on the integrity of the input walkthrough artifacts and does not perform escaping or validation of external content.
Audit Metadata