Skip to content

MilleniumGenAI/coder-openclaw-agent

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Coder for OpenClaw

coder is a coding-focused OpenClaw sub-agent for background code execution, test-driven edits, bug fixing, small project scaffolding, and small-to-medium data-analysis tasks.

ClawHub

This repository now includes a ClawHub-ready wrapper skill at clawhub/coder-openclaw-agent/.

Use that folder as the publication entrypoint for ClawHub. The root repository remains the source package for:

  • the OpenClaw prompt pack;
  • the sandbox Docker image definition;
  • the integration contract and runtime documentation.

ClawHub publishes skills under platform-wide MIT-0 terms. The repository itself keeps its own source license at the root.

For ClawHub-specific packaging and publication notes, see clawhub/README.md.

What This Project Includes

  • A hardened workspace-coder prompt pack for OpenClaw.
  • A Docker sandbox image definition for reproducible Linux-based execution.
  • A practical document/data-processing runtime for HTML, PDF, spreadsheets, and office-style files.
  • A minimal integration contract for wiring the agent into openclaw.json.

Intended Scope (v1)

Supported:

  • code execution and verification in the OpenClaw sandbox;
  • bug fixing and test-driven edits;
  • small project scaffolding;
  • small-to-medium data-analysis tasks;
  • HTML parsing and common document-processing tasks;
  • honest blocked-state reporting through PARTIAL or FAILURE.

Out of scope:

  • large monorepo refactors without tuning;
  • long-running background jobs;
  • multi-agent recursion;
  • unsupported language/toolchain stacks unless you extend the image yourself.

Requirements

  • OpenClaw 2026.3.x or later.
  • Docker available on the host.
  • An authenticated openai-codex provider profile in your OpenClaw installation.

Quickstart (10 minutes)

  1. Clone or copy this repository to any local path.
  2. Copy openclaw/workspace-coder/ into your OpenClaw base directory, or point your agent config at that path directly.
  3. Build the sandbox image:
    • docker build -f docker/coder-sandbox.dockerfile -t coder-sandbox:latest .
  4. Register the agent in openclaw.json using openclaw/agent-config.template.json as a starting point.
  5. Verify model access:
    • openclaw models status --agent coder --probe --probe-provider openai-codex --json
  6. Verify sandbox wiring:
    • openclaw sandbox explain --agent coder
  7. 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."

Required OpenClaw Workspace Files

The agent expects these injected prompt files:

  • AGENTS.md
  • SOUL.md
  • TOOLS.md
  • IDENTITY.md
  • MEMORY.md
  • USER.md
  • HEARTBEAT.md

Integration Contract

  • Runtime source of truth: openclaw/workspace-coder/SOUL.md
  • Sandbox image: coder-sandbox:latest
  • Runtime style: Linux/bash only, python3, no provider CLIs from shell
  • Working directory policy: /tmp/coder/<task_name>/
  • Output schema: strict JSON with status, task_summary, iterations_used, steps_total, steps_completed, steps_blocked, codeblocks, sandbox_log, self_corrections, and error_analysis
  • Main-agent prompt and orchestration rules: see openclaw/main-coder-prompt.md

Runtime Stack

See docker/RUNTIME.md for the preinstalled tools and Python packages included in the default image.

Validation Before Publishing

Recommended checks before publishing changes:

  • openclaw models status --agent coder --probe --probe-provider openai-codex --json
  • openclaw sandbox explain --agent coder
  • at least one successful real coding smoke task through openclaw agent --agent coder --json
  • one blocked-input task confirming honest PARTIAL or FAILURE behavior
  • ClawHub structure check: clawhub/coder-openclaw-agent/SKILL.md exists, uses YAML frontmatter, and does not require Windows-specific local paths
  • ClawHub docs smoke: a new OpenClaw user can follow clawhub/coder-openclaw-agent/SKILL.md plus this README.md without reading private memory/session files

Known Limits

See docs/known-limits.md.

Compatibility

See docs/compatibility.md.

License

MIT

About

Coding-focused OpenClaw sub-agent for background code execution, debugging, small project scaffolding, and data/document tasks.

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors