aoi-triple-memory-lite

What This Skill Does

aoi-triple-memory-lite is a lightweight, structured knowledge management system designed for OpenClaw agents. It functions as a local, file-based cognitive layer, enabling your agent to maintain long-term context without relying on cloud-based vector databases or external embeddings. By utilizing a three-tier approach—consolidated file memory, structured decision journaling, and high-speed local search—it provides a robust framework for tracking project history, evolving design choices, and research findings directly within the workspace. The skill operates entirely locally using standard Markdown and ripgrep, ensuring that your data remains portable and human-readable.

Installation

To integrate this skill into your environment, use the ClawHub command-line interface. Run the following command in your terminal:

clawhub install openclaw/skills/skills/edmonddantesj/aoi-triple-memory-lite

Ensure that you have ripgrep installed on your system path, as the search functionality relies on its performance for traversing large directories.

Use Cases

This skill is ideal for complex, multi-day development tasks where maintaining a coherent train of thought is critical. Use it to store the rationale behind architectural decisions, making it easier for the agent to revisit why a specific library or pattern was chosen weeks later. It is also excellent for ongoing research projects where you need to aggregate snippets and notes into a queryable structure without the overhead of heavy AI-agent database plugins.

Example Prompts

1. "Search my memory files for any decisions we made regarding the USDC payment gateway integration last month."
2. "Create a new decision note titled 'Authentication Provider Migration' and tag it with 'auth', 'security', and 'priority-high'."
3. "Summarize the last three entries in my memory files regarding the 'Tempo Hackathon' project and identify any open action items."

Tips & Limitations

To get the most out of aoi-triple-memory-lite, maintain consistency in your tagging strategy. Since there is no database-level schema enforcement, tagging your notes accurately is the primary way to ensure effective search results. Remember that this tool is not a synchronization engine; it operates strictly on the local workspace. If you work across multiple machines, you are responsible for keeping the memory/ and context/ directories in sync using version control (like Git). This tool is intended for lean environments and intentionally lacks automated background indexing to save system resources.