adopt
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEDYNAMIC_CONTEXT_INJECTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The skill executes a configuration helper script (
yaml-helper.sh) at load time using the!syntax to resolve project settings. - [COMMAND_EXECUTION]: The skill executes several bash scripts located within the project's
.claude/scriptsdirectory to perform format checks (gdd-structure-check.sh) and configuration migrations (migrate-v1-config.sh). It also usesmvorgit mvto reorganize files if the user agrees. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it reads and processes untrusted project documentation to generate its migration plan.
- Ingestion points: Reads files from
design/gdd/,docs/architecture/, andproduction/epics/to audit their structure. - Boundary markers: The skill does not explicitly define delimiters when processing document content for the audit report.
- Capability inventory: The agent can read files, write the migration plan, and execute shell scripts.
- Sanitization: Audit logic is partially offloaded to specific shell scripts which check for heading presence and status values.
Audit Metadata