ss-feature-workflow

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates as a thin orchestration layer, calling other vendor-owned tools (ss-create-branch, ss-proposal, ss-plan, ss-coding, ss-create-pr) to handle development tasks. It does not contain any independent code execution or direct file modification logic beyond calling these authorized sub-skills.- [PROMPT_INJECTION]: The skill provides an interface to ingest requirement documents from external URLs or local files. While this represents a potential indirect prompt injection surface, the skill includes explicit 'Iron Rules' that mandate the maintenance of safety guardrails, such as secret detection and scope-overreach rejection, even when operating in autonomous delivery modes.
  • Ingestion points: Requirement source (URL or local file) processed by the workflow.
  • Boundary markers: Relies on the execution boundaries and prompts of the child skills it orchestrates.
  • Capability inventory: Coordinates file system operations for documentation and repository management via Git/CLI tools.
  • Sanitization: Explicit instructions prevent the relaxation of safety filters and require validation of implementation requirements against the original plan.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 09:08 AM
Security Audit — agent-trust-hub — ss-feature-workflow