ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified

crew-school

Structured learning system for AI agent crews. Design curricula, run research sessions, track progress, and prevent common failure modes (lazy output, planning-without-executing). Use when setting up agent learning, running training sessions, auditing knowledge gaps, or building a curriculum for specialized agents. Works for single agents or multi-agent crews.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/crewhaus/crew-school
Or

Crew School — Agent Learning System

Run structured learning sessions that produce real knowledge articles, not plans. Prevents the #1 failure mode: agents that say "I'll research this" and stop.

Quick Start

  1. Identify gaps — What does your agent need to learn? Check what knowledge files exist vs what's needed.
  2. Pick a topic — Start with critical gaps that directly impact daily work.
  3. Spawn a session — Use the Session Template below. The anti-laziness guardrails are load-bearing — don't remove them.
  4. Verify output — Check line count, source count, and scan for planning language.

Session Template

Use this exact template when spawning learning sessions. Fill in the brackets.

CREW LEARNING SESSION — [ROLE]: [TOPIC]

You are the [ROLE] agent. Your assignment is to research [TOPIC] and produce a comprehensive knowledge article.

### Context
[2-3 sentences: why this topic matters for this agent's role]

### Prerequisites
[List knowledge files to read first, or "None"]

### Research Assignment
Deep research on [TOPIC]. You MUST cover:
1. [Subtopic 1] — [Specific questions]
2. [Subtopic 2] — [Specific questions]
3. [Subtopic 3] — [Specific questions]

### Execution Rules — READ CAREFULLY
- DO NOT PLAN. EXECUTE. If you write "I will..." or "Let me create...", STOP and DO IT instead.
- DO NOT stop after outlining. Every section must contain real research findings.
- Search the web. Perform at least 5 web searches. Read at least 2 full articles.
- Cite sources. Every claim needs a source URL.
- Be specific. Names, numbers, examples > generic advice.

### Output Requirements
Write findings to `knowledge/[filename].md` with this structure:

# [Topic]

> **TL;DR:** [2-3 sentence summary]
> **Applies to:** [Which roles]
> **Prerequisites:** [Other knowledge files]

## Key Takeaways
- [Bullet list of actionable findings]

## [Sections with real findings]

## Practical Application
[Steps, templates, checklists the agent can use immediately]

## Sources
[All URLs cited]

### Minimum Quality Thresholds
- >= 150 lines, >= 1500 words, >= 5 cited sources with URLs
- >= 4 H2 sections with substantive content
- 0 instances of "TODO", "TBD", "I will", "I'll create"
- Must include TL;DR, Key Takeaways, and Practical Application sections

### Completion
Append one line to memory/learning-log.md:
[DATE] | [ROLE] | [TOPIC] | [LINES] | [SOURCES] | [ONE-LINE SUMMARY]

Curriculum Design

For multi-session learning, create a curriculum file. See references/curriculum-design.md for:

  • How to audit knowledge gaps
  • Sequencing topics by dependency
  • Joint sessions for cross-functional learning
  • Cadence recommendations (learning vs doing ratio)

Scheduling as a Cron

Metadata

Author@crewhaus
Stars3409
Views0
Updated2026-03-25
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-crewhaus-crew-school": {
      "enabled": true,
      "auto_update": true
    }
  }
}
Safety NoteClawKit audits metadata but not runtime behavior. Use with caution.

Related Skills

crewhaus

Quick startup idea evaluation from your terminal. Score ideas on 3 dimensions, run deeper scans with real competitor data and risk assessment. A structured thinking tool for founders — not an oracle.

crewhaus 3409

action-bias

Stop agents from producing reports instead of taking action. Restructure prompts, cron jobs, and agent shifts so they DO things — send emails, post content, make API calls, push code — instead of writing plans about doing things. Use when: agents keep outputting strategy docs instead of executing, sub-agent sessions produce reports nobody reads, you notice your crew is busy but nothing external is happening, heartbeat/shift results are all 'here's what we should do' with no proof of doing it, or you want to audit whether your agents are actually shipping.

crewhaus 3409

focus-tracker

Maintain persistent focus on active work across compactions and session restarts. Use when starting a multi-step project, when resuming after compaction, when the user says "we're working on X", or when you notice you've lost track of what was being done. Also use proactively at session start to check if there's unfinished work. Prevents the agent from forgetting mid-project context, drifting between topics, or asking "what were we doing?" after a compaction.

crewhaus 3409

workspace-trash

Soft-delete protection for workspace files. Intercept file deletions and move them to a recoverable trash instead of permanent removal. Use when deleting, removing, or cleaning up files in the workspace. Triggers on /trash, /trash:view, /trash:empty, /trash:restore, or any file deletion request. Also use when the agent needs to rm/delete/remove any file — always trash instead of permanently deleting.

crewhaus 3409

recall

Teaches agents to check before they guess. Knowledge access patterns, proactive context loading, and hallucination resistance. Install before taking the Knowledge-Driven Agent certification.

crewhaus 3409