SDD Start Human Verify

AI 驱动验收执行 — 自主运行 [A] 步骤，对 [H] 步骤调用人类并提供精确的微指令。

This skill will execute shell commands found in your repository's spec files, start and stop services, update spec-human-verify.md, and may append variables into .env or modify code while attempting automatic fixes. Before installing or running it: (1) Review every spec-human-verify.md file that the skill might run to ensure it contains only safe commands. (2) Run the skill in an isolated environment (container or CI runner) with no access to secrets or production systems. (3) Backup your repo and .env before use. (4) Prefer running with human-in-the-loop (do not allow fully autonomous runs) and require explicit confirmation before any automatic edits or process kills. (5) If you plan to use it, ensure the machine has required tools (bash, ss or lsof) available and restrict its workspace_path to a safe directory.

Type: OpenClaw Skill Name: sdd-start-human-verify Version: 1.0.0 The skill 'sdd-start-human-verify' automates software acceptance testing by parsing markdown files and executing shell commands. It possesses high-risk capabilities, including the ability to execute arbitrary bash commands extracted from markdown, manage background services, and automatically modify environment variables or source code to 'fix' test failures (SKILL.md). While these behaviors are functionally aligned with the stated purpose of AI-driven verification, the broad shell execution and file-system access permissions represent a significant risk profile.