Coder for OpenClaw

Install and wire a coding-focused OpenClaw sub-agent for background code execution, test-driven edits, bug fixing, small project scaffolding, and small-to-me...

Security Scan

The skill is internally consistent with its stated purpose (wiring a coding sub-agent for OpenClaw) and asks for tools and config that match that purpose, but installing it requires you to pull and build code from a third‑party GitHub repo so you should review the repo and Dockerfile before proceeding.

Purpose & Capability

Name, description, required binaries (openclaw, docker, git), and required OpenClaw config/provider profile align with the stated goal of installing a coding-focused OpenClaw sub-agent. The listed actions (copy workspace-coder, register agent, build sandbox image) are expected for this capability.

Instruction Scope

Runtime instructions are limited to cloning the listed GitHub repo, copying agent files into your OpenClaw directory, building a Docker sandbox image, and registering the agent in openclaw.json. They do not request unrelated system files or credentials. However, the instructions do require building and running third‑party code (the Docker image and workspace contents), so you should inspect the repository and SOUL/agent files before installing.

Install Mechanism

No automated install spec is provided; installation is instruction-driven via git clone and docker build from a GitHub repo. GitHub is a common host (expected), but building a Docker image from remote source can execute arbitrary code during build — review the Dockerfile and repo before building.

Credentials

The skill requests no new environment variables itself. It does require an authenticated openai-codex provider profile in OpenClaw (expected for a coding LLM sub-agent). Be aware the sub-agent will use whatever provider credentials are present in OpenClaw (possible usage/cost implications and access to provider APIs).

Persistence & Privilege

The skill is not always:true and is user-invocable; it will register an agent in openclaw.json (expected behavior for adding a sub-agent). There is no indication it modifies other skills or system-wide agent settings beyond registering itself.

Assessment

This skill appears to do what it says, but installation requires cloning and building a third-party GitHub repo and its Docker image. Before installing: review the repository (especially Dockerfile, workspace-coder/SOUL.md, and agent-config.template.json) for unexpected actions; ensure Docker builds occur in an isolated environment if you don't fully trust the source; confirm you are comfortable granting the OpenClaw openai-codex provider profile access to this sub-agent (usage/costs and API access); back up your openclaw.json before registering the new agent; and run the provided smoke tests in a safe sandbox first.

What this skill is

This is an integration skill for installing and wiring the coder OpenClaw sub-agent from the public repository:

• coder-openclaw-agent

The repository contains:

• the workspace-coder prompt pack;
• the coder-sandbox:latest Docker image definition;
• the coder agent config template;
• the Main -> Coder orchestration contract.

This skill is intended for OpenClaw users who want a strong background coding and data-analysis sub-agent without building the orchestration from scratch.

What it can do

• code execution and verification inside the OpenClaw sandbox;
• bug fixing and test-driven edits;
• small project scaffolding;
• small-to-medium data-analysis tasks;
• HTML, PDF, spreadsheet, and office-style document processing;
• honest blocked-state reporting through PARTIAL or FAILURE.

Requirements

• OpenClaw 2026.3.x or later
• Docker available on the host
• an authenticated openai-codex provider profile

Install

1. Clone the repository:
   • git clone https://github.com/MilleniumGenAI/coder-openclaw-agent.git
2. Copy openclaw/workspace-coder/ into your OpenClaw base directory, or point your agent config at that path directly.
3. Build the sandbox image from the repository root:
   • docker build -f docker/coder-sandbox.dockerfile -t coder-sandbox:latest .
4. Register the agent in openclaw.json using:
   • openclaw/agent-config.template.json
5. If your main agent delegates coding tasks, align it with:
   • openclaw/main-coder-prompt.md

Validate

Run these checks before using the agent in real work:

openclaw models status --agent coder --probe --probe-provider openai-codex --json
openclaw sandbox explain --agent coder

Then run a first smoke task:

openclaw agent --agent coder --json --message "Return strictly valid JSON matching coder SOUL schema. GOAL: create /tmp/coder/smoke/main.py that prints hello. INPUTS: none. CONSTRAINTS: work only in /tmp/coder/smoke; use python3 and Linux/bash commands only; use PARTIAL if blocked. SUCCESS CRITERIA: python3 /tmp/coder/smoke/main.py prints hello. DELIVERABLES: codeblocks and sandbox_log."

Core references

• Root README: README.md
• Agent config template: openclaw/agent-config.template.json
• Main -> Coder orchestration guide: openclaw/main-coder-prompt.md
• Runtime inventory: docker/RUNTIME.md
• Known limits: docs/known-limits.md

Notes

• This is an OpenClaw-only v1 package.
• ClawHub publishes skills under platform-wide MIT-0 terms.
• The runtime source of truth is openclaw/workspace-coder/SOUL.md.
• Default working area inside the sandbox is /tmp/coder/<task_name>/.
• The expected output contract is strict JSON with SUCCESS | PARTIAL | FAILURE.