← Back to Skills Marketplace
295
Downloads
1
Stars
0
Active Installs
7
Versions
Install in OpenClaw
/install simple-code
Description
Plan and build small, readable coding projects in three reliable modes: start, continue, and review. Think first, make a short plan, then delegate implementa...
Usage Guidance
This skill appears coherent for small coding tasks. Before installing or invoking it, be aware that: it will read and modify files in whatever project path you point it at (or infer), will create a .steps/ folder and add .steps/ to .gitignore, and may read commits/diffs if you ask for a review. It requests no external credentials or downloads. To reduce risk, explicitly specify the project path and scope (e.g., 'review only file X' or 'start in agent_code/my-tool'), commit or back up your repo before allowing write operations, and ask the skill to run in 'review-only' mode if you want to prevent edits. If you need stronger guarantees, request that the skill produce patches instead of applying changes directly so you can review them before merging.
Capability Analysis
Type: OpenClaw Skill
Name: simple-code
Version: 1.0.2
The 'simple-code' skill bundle provides a structured framework for an AI agent to manage small coding projects using 'start', 'continue', and 'review' modes. The instructions in SKILL.md focus on standard development practices such as project organization, git usage, testing, and documentation, with no evidence of malicious intent, data exfiltration, or unauthorized execution.
Capability Assessment
Purpose & Capability
The name/description (plan, start/continue/review small projects) aligns with what the SKILL.md asks the agent to do: read/modify project files, create a minimal project structure, add tests and docs, and record progress in .steps/.gitignore. There are no unrelated env vars, binaries, or installs requested.
Instruction Scope
Instructions explicitly ask the agent to read the target project, working tree diffs, and commits and to perform file edits and testing inside agent_code/<project-name>. This is expected for a coding skill, but by default the skill may read broad parts of a project's working tree if the user doesn't restrict scope; it also creates and updates .steps/ and .gitignore.
Install Mechanism
No install spec and no code files are included (instruction-only skill), so nothing will be written to disk by an installer or downloaded during install. Runtime file I/O is limited to project work the skill performs.
Credentials
The skill requests no environment variables, credentials, or config paths. Its requested actions (reading and editing code, creating .steps) are proportionate to the stated purpose.
Persistence & Privilege
always is false and the skill does not request system-wide changes or other skills' configs. It will persist project-local artifacts (.steps/, .gitignore entries) which is expected; autonomous model invocation is allowed by default but is not requested beyond normal operation (it also suggests spawning a 'coding sub-agent' and preferring a specific model).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install simple-code - After installation, invoke the skill by name or use
/simple-code - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
Refine simple coding workflow guidance for start, continue, and review modes.
v1.2.1
Refine simple-code workflow: clarify sub-agent delegation, .steps tracking, and git/project-folder rules.
v1.2.0
**Sub-agent delegation model replaces ACP for code, test, and doc tasks.**
- Switched from ACP to using a powerful OAuth-capable coding sub-agent (preferably Codex) for implementation, testing, documentation, and tracking.
- Clarified that the chat agent keeps its earlier responsibilities: think, plan, delegate, review, and report.
- Workflow steps now specify sub-agent handles all non-planning execution in the project folder.
- Made sub-agent, not ACP, responsible for `.steps/` tracking notes and test iteration until they pass.
- Terminology and step order updated for clarity and to match the new delegation model.
v1.0.1
Initial publish after CLI update
v0.1.2
Add .steps tracking workflow for projects and commit-linked review notes; require ignoring .steps in .gitignore from project init.
v0.1.1
Require waiting for ACP Codex to finish writing before post-implementation review, testing, docs, and verification.
v0.1.0
Initial publish
Metadata
Frequently Asked Questions
What is Simple Code?
Plan and build small, readable coding projects in three reliable modes: start, continue, and review. Think first, make a short plan, then delegate implementa... It is an AI Agent Skill for Claude Code / OpenClaw, with 295 downloads so far.
How do I install Simple Code?
Run "/install simple-code" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Simple Code free?
Yes, Simple Code is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Simple Code support?
Simple Code is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Simple Code?
It is built and maintained by Yi (@hyharry); the current version is v1.0.2.
More Skills