A local, consent-based memory skill that allows AI agents to adapt to your specific coding styles and stack preferences through explicit feedback.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install coding-custom
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install coding-custom using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Coding Style Memory?
The Coding skill is a sophisticated memory layer designed for developers who want their AI agents to maintain consistency across different sessions. Unlike traditional agents that might hallucinate or forget your specific formatting rules, this tool creates a tiered local storage system that records your explicit corrections and confirmations. It acts as a bridge between your personal dev standards and the AI's output generation.
Built as part of the Openclaw Skills ecosystem, it prioritizes user sovereignty and privacy. The skill never scans your private repositories or observes your behavior silently; instead, it waits for explicit instructions to learn, ensuring that the memory stays clean, relevant, and strictly controlled by the developer.
Coding Style Memory Use Cases
- Enforcing specific linting or naming conventions (e.g., snake_case vs camelCase) across all AI-generated files.
- Maintaining a consistent tech stack and library versions to prevent the AI from suggesting outdated or irrelevant dependencies.
- Storing organizational architectural patterns, such as preferred folder structures or testing strategies.
- Archiving rejected patterns to ensure the agent never suggests previously discarded solutions.
How Coding Style Memory Works
- At the start of a session, the agent loads existing preferences from ~/coding/memory.md.
- When the user corrects the AI's output, the skill triggers an inquiry to see if the preference should be permanent.
- Upon user confirmation, the preference is distilled into an ultra-compact format (5 words or less).
- New entries are categorized into Stack, Style, Structure, or Never categories within the local memory file.
- The skill monitors file size, automatically archiving older entries to history.md when the active memory exceeds 100 lines to prevent context bloat.
Coding Style Memory Setup
To begin using this skill with your agent, ensure the local directory exists:
mkdir -p ~/coding
You can manage the skill lifecycle using standard Openclaw Skills commands:
clawhub sync
# To support the project
clawhub star coding
Coding Style Memory Data Schema & Taxonomy
The skill organizes its memory within the ~/coding/ directory using a tiered markdown structure:
| File | Purpose | Limit |
|---|---|---|
memory.md |
Active, high-priority preferences used in context. | Max 100 lines |
history.md |
Archived preferences and old patterns. | Unlimited |
criteria.md |
Definitions for when a preference deserves to be saved. | N/A |
dimensions.md |
Taxonomy for preference categories (Stack, Style, etc.). | N/A |
Entry Format:
Preferences are stored as brief, colon-separated strings such as python: prefer 3.11+ or naming: snake_case.
Coding Style Memory Advanced Features
- Automatic Archiving: Intelligently moves less frequent preferences to history.md to maintain high performance and low context costs.
- Context Injection: Seamlessly applies stored rules to every code generation prompt without manual user intervention.
- Multi-Agent Support: Because the data is stored in standard Markdown files in a predictable local path, multiple Openclaw Skills can interface with the same preference set.
- Semantic Consolidation: Capability to merge similar entries (e.g., merging multiple 'no-tooling' preferences) to save space.
SKILL.md
Loading
Related Openclaw Skills
Book Highlights
A specialized AI agent skill designed to extract and format the most impactful highlights, quotes, and core insights from an extensive library of books.
CrewAI
CrewAI is a sophisticated multi-agent orchestration framework designed to help developers build collaborative AI systems where specialized agents work together to solve complex tasks.
Zhuihui Branch Attendance Analyzer
An automated analysis tool for processing complex employee attendance records and generating professional docx and xlsx reports.
DeFi Scout
DeFi Scout delivers real-time on-chain financial intelligence, covering yields, prices, and protocol health across the Optimism and Base ecosystems.
GraphCare
A zero-trust structural database scanner that audits schema topology for performance bottlenecks and integrity gaps without accessing row data.
Knowledge Capture
An intelligent knowledge management and adaptive review system that converts conversations into structured Markdown notes and automated study plans.
Featured*
