New to Claude Skills? Learn how to install them →

Get Claude Skills
ruvnet on GitHub

Neural Training

Free

Train and consolidate neural patterns efficiently.

by ruvnet67.6k stars on ruvnet/ruflo
Updated Aug 10, 2026
Get this skill

Free · Opens the source repo

What Neural Training does

Neural Training is a specialized skill designed for training and consolidating neural patterns based on successful task completions. It implements crucial phases of the intelligence pipeline, specifically the DISTILL and CONSOLIDATE phases, allowing users to effectively capture and store learned behaviors. This skill is particularly useful for developers and AI practitioners who need to refine their models based on real-world task performance.

The DISTILL phase begins by checking the current neural status and initiating a trajectory that records significant actions during task execution. Each action's outcome is assessed, and the system learns from this trajectory to improve future performance. By training patterns and storing them, users can ensure that their models adapt and evolve based on their experiences. This process is essential for maintaining an up-to-date and relevant AI model that reflects the latest successful strategies.

The CONSOLIDATE phase complements the training by folding learned patterns into long-term storage, preventing the loss of valuable knowledge through mechanisms like EWC++. This is particularly important when dealing with multiple domains, as users can create MicroLoRA adapters to manage domain-specific knowledge without overwriting existing patterns. This skill is ideal for those looking to enhance their AI systems' capabilities through structured learning and adaptation processes.

When to use it

Use this skill after completing successful tasks to capture effective strategies, or when training in new domains to create tailored adaptations.

When not to use it

This skill may not be suitable for scenarios where real-time adaptation is not required, or when working with a single domain that does not necessitate pattern consolidation.

What you can build with it

Post-Task Learning

After completing a task, use Neural Training to capture what strategies worked effectively.

Domain-Specific Adaptation

When entering a new domain, create a MicroLoRA adapter to tailor your AI's responses.

Long-Term Knowledge Management

After every ten task completions, run the CONSOLIDATE phase to ensure your AI retains valuable patterns.

How to install Neural Training

View source

1. Install with the skills CLI

npx skills add ruvnet/ruflo/neural-train --agent claude-code

2. Or install it manually

Download the skill folder and drop it into ~/.claude/skills/ for all projects, or .claude/skills/ to scope it to one repo. Restart Claude Code so it picks up the new skill.

Anthropic's agentic coding CLI, and the reference implementation of Agent Skills. Drop a skill folder into ~/.claude/skills and Claude Code loads it automatically whenever a task matches the skill's description. Claude Code docs

Inside SKILL.md

Written by ruvnet

Neural Training

Train and consolidate neural patterns. Implements the DISTILL and CONSOLIDATE phases of the 4-step intelligence pipeline.

When to use

  • After completing a successful task — capture what worked.
  • After accumulating ≥10 task completions — run consolidation to fold patterns into long-term storage.
  • When training a new domain — create a MicroLoRA adapter for it.

Standard flow (DISTILL)

  1. Check current neural status — mcp__plugin_ruflo-core_ruflo__neural_status.
  2. Start a trajectory — mcp__plugin_ruflo-core_ruflo__hooks_intelligence_trajectory-start with the task context.
  3. Record steps — for each significant action, mcp__plugin_ruflo-core_ruflo__hooks_intelligence_trajectory-step.
  4. End trajectory — mcp__plugin_ruflo-core_ruflo__hooks_intelligence_trajectory-end with verdict: pass|fail|partial.
  5. Learn from the trajectory — mcp__plugin_ruflo-core_ruflo__hooks_intelligence_learn.
  6. Train patterns — mcp__plugin_ruflo-core_ruflo__neural_train with --pattern-type coordination --epochs 10.
  7. Store patterns — mcp__plugin_ruflo-core_ruflo__hooks_intelligence_pattern-store.
  8. Verify — mcp__plugin_ruflo-core_ruflo__neural_patterns to confirm.

SONA adaptation (single-domain, <0.05ms)

For real-time micro-adaptation:

mcp tool call ruvllm_sona_create --json -- '{"domain": "coding"}'
mcp tool call ruvllm_sona_adapt --json -- '{"feedback": {"score": 0.9, "trajectory": "..."}}'

MicroLoRA adaptation (multi-domain)

When you have ≥3 distinct domains, create a MicroLoRA adapter per domain rather than overloading SONA:

# Create the adapter
mcp tool call ruvllm_microlora_create --json -- '{"domain": "frontend"}'

# Adapt with feedback
mcp tool call ruvllm_microlora_adapt --json -- '{"adapter": "frontend", "feedback": {...}}'

# CONSOLIDATE phase: apply EWC++ on weight deltas to prevent catastrophic forgetting
mcp tool call ruvllm_microlora_adapt --json -- '{"adapter": "frontend", "consolidate": true}'

The --consolidate flag is the EWC++ trigger. Without it, fresh training overwrites older domains.

CONSOLIDATE phase (separate from training)

After every ~10 trajectory completions, run a full consolidation pass:

mcp tool call agentdb_consolidate --json
mcp tool call neural_compress --json    # storage efficiency

This folds patterns into long-term storage under EWC++ semantics.

Bootstrapping from scratch

If the system has no learned patterns yet:

mcp tool call hooks_pretrain --json -- '{"modelType": "moe", "epochs": 10}'
mcp tool call hooks_build-agents --json -- '{"agentTypes": "coder,tester"}'

hooks_pretrain writes to the patterns (plural) namespace — distinct from the pattern (singular) ReasoningBank target. See ruflo-agentdb ADR-0001 for the namespace convention.

Reset (testing only)

To wipe intelligence state (e.g., for benchmarking):

mcp tool call hooks_intelligence-reset --json

CLI alternatives

npx @claude-flow/cli@latest neural train --pattern-type coordination --epochs 10
npx @claude-flow/cli@latest neural patterns --list
npx @claude-flow/cli@latest neural status
npx @claude-flow/cli@latest neural compress
npx @claude-flow/cli@latest hooks pretrain --model-type moe --epochs 10
npx @claude-flow/cli@latest hooks build-agents --agent-types coder,tester

Frequently asked questions about Neural Training

Similar skills

Skill Creator

Efficiently create and manage skills for Gemini CLI.

Agent Development

Create and manage autonomous agents for Claude Code.

Math Olympiad Solver

Solve and verify competition math problems effectively.

Microsoft Skill Creator

Create specialized skills for Microsoft technologies.

Doublecheck

A verification pipeline for AI-generated claims.

Skill Development for Claude Code

Create and enhance skills for Claude Code plugins.