项目开发助理

Project Development Assistant — structured logging + issue tracking with anti-spam design. Trigger ONLY when user explicitly says new project, continue proje...

What this skill will do: it scans your filesystem for project folders (home workspace and common drive roots), creates a standard project folder structure when asked, writes daily devlog files and a structured .project_state.json, and stores briefings under the skill's briefings directory. It does not require any API keys or network installs. Things to consider before installing: 1) Filesystem access: the code will list directories on C:/, D:/, E:/, F: and ~/.openclaw/workspace to find projects — ensure you are comfortable with that scope and the skill having read access to those locations. 2) Writes and deletions: it will create folders and files in the project path you supply and in the skill's 'briefings' directory; project_admin also removes briefing files older than 3 days and can purge test-named briefings. Back up any important files first. 3) Mismatch note: SKILL.md mentions internet searches for component info, but none of the supplied scripts perform network requests — if you need automatic web lookups, verify how and whether that will be added. 4) Activation: SKILL.md states the skill should only trigger on explicit commands (new project / continue / record etc.). Confirm the agent's invocation policy respects that. If you want stronger safety, run the scripts in a sandboxed environment or inspect/run them manually before enabling autonomous invocation.

Type: OpenClaw Skill Name: project-development-assistant Version: 1.0.0 The 'project-development-assistant' skill bundle is a structured project management tool designed to help AI agents track development progress, log activities, and manage session context. The Python scripts (log_project.py, task_briefing.py, etc.) focus on file-based state management, directory structure visualization, and log analysis within user-specified project paths and common workspace locations (e.g., ~/.openclaw/workspace). While the scripts scan local drives (C:, D:, E:, F:) to identify project folders containing a 'logs' directory, there is no evidence of data exfiltration, unauthorized network activity, credential theft, or malicious prompt injection. The behavior is consistent with the stated purpose of maintaining project continuity across multiple sessions.