skills/steveclarke/dotfiles/devstack/Gen Agent Trust Hub

devstack

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates and executes local shell scripts (bin/dev, bin/worktree) to orchestrate development processes and manage git worktrees. These scripts are created from documented templates provided in the skill files.
  • [EXTERNAL_DOWNLOADS]: Recommends installing process-compose and outport via Homebrew as functional dependencies. These tools are required for the orchestration features described in the documentation.
  • [DATA_EXFILTRATION]: The worktree management script uses pbcopy to place configuration summaries on the user's system clipboard. This is a documented feature intended to assist users in transferring context between agent sessions.
  • [PROMPT_INJECTION]: Analyzed the surface for indirect prompt injection via GitHub issue titles. * Ingestion points: Issue titles fetched via gh issue view in bin/worktree. * Boundary markers: None present in the raw input. * Capability inventory: Directory creation (mkdir), git worktree operations (git worktree add), and port allocation (outport up). * Sanitization: Implements a robust slugification function using tr, sed, and awk that filters all characters except lowercase alphanumeric characters and dashes, effectively neutralizing injection attempts.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 12:52 PM