x9-decision-map
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation and management of decision maps by reading from and writing to Markdown files in the local repository (e.g.,
docs/decision-maps/). This behavior introduces a surface for indirect prompt injection if the processed files or referenced evidence contain malicious instructions. - Ingestion points: The agent reads existing initiative maps and evidence records (which may include external URLs, paths, or conversation snippets) from the local filesystem.
- Boundary markers: The instructions do not define specific delimiters or warnings to prevent the agent from executing instructions potentially embedded within the decision maps or evidence text.
- Capability inventory: The skill possesses the authority to read, create, and modify local files, and it can hand off data to other engineering skills (
ce-brainstorm,ce-plan). - Sanitization: There is no evidence of content sanitization or validation performed on the ingested Markdown data before it is re-integrated into the agent's context or written back to the disk.
Audit Metadata