← 返回 Skills 市场
449
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install preqstation-openclaw
功能描述
Delegate PREQSTATION coding tasks to Claude Code, Codex CLI, or Gemini CLI with PTY-safe execution (workdir + background + monitoring). Use when building, re...
安全使用建议
This skill is coherent for delegating coding tasks to local CLI engines, but before installing: (1) ensure you have and trust the claude/codex/gemini CLI binaries the skill will invoke, (2) be aware it will read and update MEMORY.md in the skill repo to store absolute paths you provide and will run git/worktree commands and engine commands inside those mapped project worktrees, so only add project paths you control (avoid mapping system or sensitive directories), (3) confirm the OPENCLAW_WORKTREE_ROOT location (default /tmp/openclaw-worktrees) is acceptable, and (4) if you need more restriction, do not add mappings or grant the agent autonomous runs that will launch engine commands without explicit confirmation.
功能分析
Type: OpenClaw Skill
Name: preqstation-openclaw
Version: 0.1.6
This skill is classified as suspicious primarily due to its explicit instructions to disable sandboxing for the underlying coding agents (Claude, Codex, Gemini) using flags like `--dangerously-skip-permissions`, `--dangerously-bypass-approvals-and-sandbox`, and `GEMINI_SANDBOX=false` in SKILL.md. While the skill attempts to mitigate risk through worktree isolation and other 'safety gates' (also defined in SKILL.md), disabling these critical security features creates a significant vulnerability. Additionally, the skill instructs the agent to clone arbitrary user-specified repositories for PR review and to read/write user-provided absolute paths to MEMORY.md, further increasing the attack surface if the agent's internal instructions are bypassed or misinterpreted.
能力评估
Purpose & Capability
The skill claims to delegate coding tasks to local CLI engines (claude/codex/gemini) and to manage git worktrees; its declared prerequisites (git and at least one engine binary) and use of a per-task worktree are appropriate and proportional to that purpose.
Instruction Scope
Instructions direct the agent to read and update the skill's MEMORY.md (a Projects table of absolute paths) and to create and operate inside git worktrees derived from those paths. This is coherent with the stated goal, but it means the skill will be given access to arbitrary absolute project paths you provide and will run engine commands in those worktrees (i.e., it can execute arbitrary operations inside mapped project directories).
Install Mechanism
There is no install spec and no code files to install or execute; the skill is instruction-only, which minimizes supply-chain/install risk.
Credentials
The skill requests no secrets and only optionally uses OPENCLAW_WORKTREE_ROOT. That is appropriate. However, it requires the agent to resolve absolute filesystem paths (project_cwd) provided by the user and will create worktrees under a chosen root, so filesystem access is inherent and must be considered by the user (do not map sensitive system directories).
Persistence & Privilege
always:false and no special system config or other skills' config are requested. The skill will update its own MEMORY.md to store project mappings — this is expected behavior and limited to its repository file.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install preqstation-openclaw - 安装完成后，直接呼叫该 Skill 的名称或使用
/preqstation-openclaw触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v0.1.6
preqstation-openclaw 0.1.6
- Major update: comprehensive SKILL.md written to define task handling, project mapping, execution flows, and hard safety requirements.
- Introduces detailed rules for project key/path resolution using MEMORY.md and user interaction when mappings are missing.
- Defines strict execution isolation: all coding tasks run in per-project git worktrees, never in live checkouts.
- Clearly documents bash interface and PTY-based engine invocation for Claude, Codex, and Gemini CLI.
- Enforces project branch/worktree naming conventions and absolute path handling.
- Details runtime prerequisites, environment variables, and strict safety gates for all engine commands.
v0.1.0
Initial release of preqstation-openclaw: Trigger PREQSTATION task execution via messenger commands with project-based routing and engine choice.
- Parses messenger command text for task triggers.
- Requires explicit selection of execution engine (`claude` or `codex`).
- Resolves project workspace paths via ticket prefix mapping.
- Executes in "dangerous mode" with a warning before execution.
- Fails fast if PREQSTATION MCP is not set up.
- Does not store API credentials; operates via MCP tools.
- Project mappings and guardrails managed in `memory.md`.
元数据
常见问题
Preqstation 是什么？
Delegate PREQSTATION coding tasks to Claude Code, Codex CLI, or Gemini CLI with PTY-safe execution (workdir + background + monitoring). Use when building, re... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 449 次。
如何安装 Preqstation？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install preqstation-openclaw」即可一键安装，无需额外配置。
Preqstation 是免费的吗？
是的，Preqstation 完全免费（开源免费），可自由下载、安装和使用。
Preqstation 支持哪些平台？
Preqstation 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Preqstation？
由 sonim1（@sonim1）开发并维护，当前版本 v0.1.6。
推荐 Skills