get-shit-done

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection vulnerability surface.
  • Ingestion points: The skill is instructed to load files referenced by @path as mandatory context and to load sub-skills based on user-provided command prefixes.
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard embedded instructions within the ingested content.
  • Capability inventory: The skill utilizes the Bash tool, Read/Write tools, and subagent execution.
  • Sanitization: No sanitization or validation of the external file content or command strings is described.
  • [COMMAND_EXECUTION]: Dynamic loading of instructions and agents from computed paths.
  • The skill resolves user commands to file paths using the pattern {{PLATFORM_ROOT}}/skills/gsd-/SKILL.md and identifies subagents within {{PLATFORM_ROOT}}/agents.
  • This dynamic loading is the core mechanism for the skill's extensible architecture but requires that the local skill directory remains secure from unauthorized modifications.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 08:17 PM