test-spec-writer
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts (
trace_lint.pyandtrace_build_rtm.py) within theplugins/testany-eng/directory to perform validation and traceability checks. These are vendor-specific tools relevant to the skill's purpose. - [INDIRECT_PROMPT_INJECTION]: The skill uses
Globto scan various document types (PRD, API, HLD, LLD, etc.), creating an ingestion point for untrusted data that could contain malicious instructions. - Ingestion points: Project documents scanned by the
Globtool in Phase 0 (PRD, API, HLD, LLD, Test Strategy, Guardrails). - Boundary markers: Includes metadata delimiters (
TRACEABILITY-METADATA:BEGIN/END) but lacks explicit instructions to ignore instructions found within the scanned documents. - Capability inventory: Execution of local Python scripts, file system scanning (
Glob), and task tracking tools. - Sanitization: No evidence of sanitization or filtering of the content extracted from external documents is present in the instructions.
Audit Metadata