项目制多智能体开发协议
PassAudited by ClawScan on May 1, 2026.
Overview
This skill is a disclosed project-workflow helper, but users should notice that it keeps persistent project files and references local helper scripts before using it.
This appears safe to install for project workflow use. Before using it, review the local SOP and dev_Project/dev_project files it will rely on, confirm the task breakdown before allowing later phases, and do not run the optional memory_viewer.py command unless you have reviewed that local script.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If those local files contain sensitive data or are modified incorrectly, future project-mode runs may rely on stale, private, or misleading context.
The skill uses local persistent project/protocol files as context and updates the project file during the workflow.
读取 `system_protocol_project_mode.md` 了解 SOP 读取 `dev_project.md` 获取项目清单 将拆解清单更新到 dev_project.md
Review the SOP and dev_Project/dev_project files before use, avoid storing secrets in them, and keep them under user control.
The agent may proceed through a multi-step workflow after activation, so the user should monitor scope and progress.
The skill explicitly describes background task decomposition, role dispatch, and retry handling.
该工具将在后台自动拆解任务、调度程序员和测试员、更新 dev_project.md 并处理错误重试。
Use the skill for clearly bounded projects, verify the architecture plan before continuing, and stop the workflow if it begins changing files outside the intended project.
If the user runs that helper, its behavior depends on a local file outside the reviewed artifacts.
The documentation references running a local helper script that is not included in the provided manifest or install specification.
python3 ~/.openclaw/workspace/memory_viewer.py
Inspect ~/.openclaw/workspace/memory_viewer.py before running it, and only execute it if it comes from a trusted source.