Workflow Crystallizer for Openclaw

An intelligent pattern-mining tool that transforms recurring memory logs into actionable automation suggestions like cron jobs and skill drafts.

newageinvestments25-byte
v1.0.0
Mar 30, 2026
0
744
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install workflow-crystallizer

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install workflow-crystallizer 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 Workflow Crystallizer?

Workflow Crystallizer is a specialized tool within the Openclaw Skills ecosystem designed to analyze agent memory logs for recurring patterns, repeated requests, and multi-step workflows. Unlike standard log analyzers, this skill maintains a persistent state, remembering previous suggestions and user feedback to ensure that only high-confidence, novel insights are presented.

By mining historical data stored in daily markdown files, the Workflow Crystallizer identifies opportunities for optimization, such as creating new workflow shortcuts or scheduling recurring tasks. This allows the AI agent to essentially build its own efficiency tools over time, evolving alongside the user's specific habits and requirements.

Workflow Crystallizer Use Cases

  • Identifying recurring manual requests that occur more than three times over several days.
  • Formalizing complex, multi-step sequences into dedicated skill drafts for better execution.
  • Spotting time-correlated tasks that are perfect candidates for automated cron job definitions.
  • Reducing cognitive load by automating repetitive monitoring or reporting tasks discovered in logs.

How Workflow Crystallizer Works

  1. The system scans the memory directory to ingest daily log files for analysis.
  2. Pattern detection scripts cluster events based on similarity, frequency, and temporal correlation.
  3. Suggestions are filtered through a confidence threshold (default 60%) to ensure high-quality proposals.
  4. The tool cross-references existing skills and crons to avoid redundant automation suggestions.
  5. New insights are compiled into a report while the internal state is updated to track accepted, rejected, or snoozed items.

Workflow Crystallizer Setup

To get started with this component of your Openclaw Skills library, navigate to the skill directory and execute the analysis pipeline:

cd /Users/openclaw/.openclaw/workspace/skills/workflow-crystallizer
# Initial full run to build state
python3 scripts/analyze_patterns.py --full | python3 scripts/generate_suggestions.py | python3 scripts/report.py --state-file state.json

For ongoing maintenance, use the incremental run command to process only new logs:

python3 scripts/analyze_patterns.py | python3 scripts/generate_suggestions.py | python3 scripts/report.py --state-file state.json

Workflow Crystallizer Data Schema & Taxonomy

The skill organizes its intelligence using the following structure:

File/Directory Purpose
memory/*.md Source data: Daily logs analyzed for pattern detection.
state.json Persistent storage: Tracks cached events, processed dates, and suggestion status.
scripts/ Logic layer: Python scripts for analysis, suggestion generation, and reporting.
references/ Documentation: Defines scoring weights and pattern detection logic.

Workflow Crystallizer Advanced Features

  • Incremental processing to minimize token usage by only analyzing new memory files.
  • Sophisticated suggestion lifecycle management including snooze functionality and rejection persistence.
  • Configurable confidence thresholds and evidence requirements for different pattern types.
  • Integration-ready reporting that can be scheduled via weekly heartbeat or cron jobs for consistent workflow optimization.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*