fsxmemory

What This Skill Does

Foresigxt Memory (fsxmemory) is a specialized, structured memory management system designed specifically for AI agents operating within the OpenClaw ecosystem. It acts as an persistent external 'long-term memory' layer that solves the common issue of context window exhaustion or 'context death.' Unlike standard text logs, fsxmemory categorizes information into semantic buckets such as facts, decisions, lessons, commitments, and projects. It is fully compatible with Obsidian markdown, allowing users to browse their agent's insights in any standard markdown editor. By providing mechanisms for checkpoints, recaps, and handoffs, it enables agents to maintain continuity across long-running tasks or multi-day projects, ensuring that vital context is preserved between session terminations.

Installation

To begin, ensure you have npm installed. Install the CLI tool globally using: npm install -g @foresigxt/foresigxt-cli-memory. Once installed, initialize your memory vault by running fsxmemory init ~/memory. You can configure the vault location globally using the FSXMEMORY_PATH environment variable or define it on a per-project basis by placing a .env file in your workspace root.

Use Cases

This skill is essential for complex, multi-step engineering tasks where an agent needs to remember previous design decisions or procedural knowledge. It is ideal for developers who need to track project blockers across different days, teams working on shared research that requires persistent semantic memory, and users who want a searchable, human-readable record of their AI interactions that doesn't disappear when a chat session ends.

Example Prompts

1. "fsxmemory remember decision 'Switching to PostgreSQL' --content 'Required for ACID compliance in our transaction logs.'"
2. "fsxmemory handoff --working-on 'API Integration' --blocked 'Missing documentation for auth endpoints' --next 'Finish the request payload parser.'"
3. "fsxmemory vsearch 'What are the main technical blockers preventing our deployment?'"

Tips & Limitations

To get the most out of fsxmemory, always perform a checkpoint before closing a session, especially if you are working on complex code refactoring. Use the recap command at the start of every day to ground the agent in its current state. Note that semantic search (vsearch) performs significantly better when your memory entries are descriptive. Be aware that this skill performs local file system operations; ensure your memory vault is backed up if it contains sensitive project information, as local file-read/write permissions are required for the agent to function.