skill-from-memory

What This Skill Does

The skill-from-memory agent is a powerful automation utility designed to bridge the gap between ad-hoc problem solving and permanent, shareable tooling. It allows users to capture the context of successful workflows, code snippets, and decision-making logic directly from conversation histories or reflective memory files. By automating the extraction and packaging process, this skill ensures that valuable insights are not lost to the "infinite scroll" of history, but instead become modular, reusable OpenClaw skills that can be pushed to GitHub or published to the ClawHub registry for community use.

Installation

You can install this skill directly via the ClawHub registry to begin transforming your history into actionable assets: clawhub install openclaw/skills/skills/zfanmy/skill-from-memory Ensure you have the OpenClaw CLI configured with your GitHub and ClawHub credentials before attempting to execute the publish.sh script.

Use Cases

• Automating Repetitive Tasks: If you spend an hour solving a unique API integration, use this skill to capture the logic and turn it into a reusable tool.
• Knowledge Management: Convert detailed research sessions or brainstorming logs into structured reference skills for future team onboarding.
• Open Source Contribution: Quickly package "aha!" moments from your development cycles into clean, documented skills for the wider OpenClaw ecosystem.

Example Prompts

1. "OpenClaw, look at my conversation history from last Thursday and identify the logic I used for the database backup; package it into a new skill called 'automated-db-backup'."
2. "Extract the steps I just followed to sanitize the API logs from my memory file for today and create a reusable 'log-scrubber' tool."
3. "Analyze my last project session, create a skill structure from the code I wrote, and publish it to the clawhub under the slug 'data-formatter-pro'."

Tips & Limitations

• Review Before Publishing: Because this skill automates code and script generation, always audit the output in the ./scripts/ directory before running the final publish.sh command.
• Clean Inputs: Providing focused context using the --since and --pattern flags in extract-from-history.sh results in much cleaner, less cluttered skill structures.
• Security: Be cautious when using this on conversations containing sensitive API keys or credentials; manually sanitize generated references/ or assets/ files to ensure no secrets are published to public repositories.