soulflow
General-purpose AI workflow framework for OpenClaw. Build custom multi-step workflows for any task — dev, ops, research, content, or automation. Ships with dev workflow examples.
53
Does it follow best practices?
If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:
npx tessl skill review --optimize ./path/to/skillValidation for skill structure
Evaluation results
100%
67%Automate the Release Cycle with a Custom Workflow
Custom workflow JSON structure
Top-level id field
0%
100%
Top-level name field
100%
100%
Top-level version field
0%
100%
Top-level description field
100%
100%
Steps array present
100%
100%
Step id fields
100%
100%
Step name fields
100%
100%
Step input fields
0%
100%
Step expects fields
0%
100%
Step maxRetries fields
0%
100%
{{task}} variable usage
0%
100%
Explicit tool instructions
0%
100%
Anti-vague instruction
0%
100%
Workflow file location
44%
100%
Without context: $0.6939 · 2m 19s · 34 turns · 36 in / 8,231 out tokens
With context: $0.5178 · 2m 3s · 23 turns · 973 in / 6,771 out tokens
90%
38%Build a SoulFlow Router for the Team Assistant
Workflow invocation and keyword routing
Correct base directory
0%
100%
Uses node soulflow.js run
0%
100%
Security audit keywords
75%
100%
Bug fix keywords
100%
100%
Feature dev keywords
41%
100%
Unmatched request handling
83%
100%
Custom workflow check
0%
16%
Task string passed to run
100%
100%
Correct workflow names
60%
100%
Without context: $0.5706 · 1m 37s · 31 turns · 9,609 in / 5,488 out tokens
With context: $0.3470 · 1m 8s · 17 turns · 23 in / 3,804 out tokens
100%
92%Research Pipeline for Competitive Analysis
Variable extraction and step retry logic
Uses {{stepid_output}}
0%
100%
KEY: value output extraction
0%
100%
Extracted variable used downstream
0%
100%
onFail retry configured
0%
100%
STATUS: done in all steps
0%
100%
Explicit tool instructions
0%
100%
Anti-vague instruction
0%
100%
Step focus
100%
100%
Workflow file extension
0%
100%
All required step fields
0%
100%
Without context: $0.8933 · 3m 3s · 49 turns · 1,023 in / 10,294 out tokens
With context: $0.4358 · 1m 36s · 21 turns · 1,297 in / 5,205 out tokens
- Evaluated
- Agent
- Claude Code
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.