gsd-discuss-phase
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
gsd-sdkutility via theBashtool to retrieve local configuration settings. - [DYNAMIC_EXECUTION]: It dynamically loads and follows instructions from markdown files located at
$HOME/.claude/get-shit-done/workflows/depending on the--assumptionsflag or theDISCUSS_MODEsetting. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project files which may contain hidden prompt injection attacks.
- Ingestion points: Reads
PROJECT.md,REQUIREMENTS.md,STATE.md, and priorCONTEXT.mdfiles for context gathering. - Boundary markers: No delimiters or explicit instructions are used to isolate untrusted content from the agent's internal logic.
- Capability inventory: The agent has permissions to execute shell commands (
Bash), modify the filesystem (Write), and invoke sub-agents (Agent). - Sanitization: No evidence of content sanitization or validation was found for the project data being processed.
Audit Metadata