teach

name	teach
description	IRIS Teach skill. USE WHEN the user wants to learn about a topic or generate educational material. Guides users through defining a topic, configuring learning parameters (level, audience, style), and launching 6 specialized pedagogical agents to create adaptive educational content. Independent from the research router.
tools	Read, Write, Edit, Bash, Grep, Glob, Task

IRIS Teach

Educational content generation skill. This is the domain-specific skill for creating learning materials on ANY topic: history, geography, science, mathematics, philosophy, arts, languages, technology, health, and any other subject. It guides users through topic definition, learning parameter configuration, and launches an orchestrated investigation with 6 specialized pedagogical agents.

Workflow Routing

Workflow	Trigger	File
Investigate	"teach", "learn", "aprender", "educate"	workflows/investigate.md
Orchestrate	(internal — invoked by investigate)	workflows/orchestrate.md
Continue	"continue", "deepen"	../research/workflows/continue.md (shared)
Status	"teach status", "check learning"	../research/workflows/status.md (shared)
Archive	"archive teach material"	../research/workflows/archive.md (shared)

User Communication Guide

The teach domain targets non-technical users (students, educators, general public). All user-facing messages MUST use humanized language. Internal processing, variables, file paths, and metadata continue to use technical terms.

Why: A student asking "teach me about the French Revolution" should never see words like "agent", "orchestrate", "spawn", or "workflow". These are implementation details that would confuse or alienate non-technical users.

Rule: Use humanized terms ONLY in ask_user messages and direct responses to the user. Variables, YAML keys, file paths, code, and internal logic ALWAYS use technical terms (agent, spawn, orchestrate, workflow).

Internal Term	User-Facing Term	Example
agent	specialist	"6 specialists will work on your topic"
spawn agents	put specialists to work	"I'll put the team to work now"
agents completed	specialists finished	"4 specialists have finished their work"
agents still running	specialists still working	"2 specialists are still working"
agent failed	specialist couldn't complete	"One specialist couldn't finish their task"
orchestrate	coordinate	Never shown to user directly
workflow	(invisible)	Never mentioned to user
synthesizer	(part of "the team")	Never named individually to user
wait gate	waiting for the team	"Waiting for all specialists to finish"
request file	learning session	"I found a previous learning session"
investigation	learning session	"Starting your learning session"
research run	session run	Internal only

How It Works

1. User requests to learn about a topic
2. Investigate workflow activates:
   • Checks for existing investigations (offers to continue or start new)
   • Gathers topic, education level, target audience, learning style
   • Confirms understanding with the user
   • Generates a request file in .requests/ with GUID-based ID
   • Creates a UUID run folder in .research/{uuid}/agents/
   • Transitions to the orchestration phase
3. Orchestration phase:
   • Puts 6 specialists to work in parallel
   • Monitors completion (mandatory wait — never skips anyone)
   • Triggers the final synthesis
   • Status tracked in request file (researching → synthesizing → completed)
4. Results appear in .research/{uuid}/SUMMARY.md as a complete educational document

Specialist Team (6)

Agent	Role	Specialization
teacher-researcher	Investigator	Gathers information on any topic
teacher-curriculum-designer	Curriculum architect	Structures learning path with prerequisites and progression
teacher-content-creator	Content producer	Creates explanations, analogies, examples, visual aids
teacher-assessor	Evaluator	Designs exercises, quizzes, Socratic questions, rubrics
teacher-fact-checker	Credibility guardian	Validates sources, experts, detects biases, flags misinformation
teacher-synthesizer	Synthesizer	Combines all outputs into the final educational document

Educational Parameters

Parameter	Values	Description
education-level	basic, intermediate, advanced	Complexity depth
target-audience	children, teenagers, adults, professionals	Tone and vocabulary adaptation
learning-style	visual, textual, practical, mixed	Content presentation style
include-exercises	yes, no	Whether to include assessment materials
language	any language	Language for the final document

Cumulative Research

Educational investigations support multiple runs for iterative deepening:

• Each run creates a new UUID folder under .research/
• The request file tracks all runs in learning-runs
• Use /iris:research continue to deepen an existing investigation
• The continue workflow reads previous SUMMARY.md and injects context into new agents

Principles

Conversational Refinement

Never assume you understand the topic. Always:

1. Restate what you understood
2. Ask for confirmation using ask_user
3. Only proceed after explicit user approval

Progressive Disclosure

Start broad, then refine:

1. What do you want to learn about?
2. What is your current knowledge level?
3. Who is the target audience?
4. What specific aspects interest you most?
5. What language should the material be in?

Educational Quality

• Every factual claim must have an identifiable source
• Controversial topics include multiple perspectives
• Content adapts to the specified education level and audience
• Assessment materials align with learning objectives

Request File Contract

Every investigation produces a request file in .requests/ with:

• Full frontmatter (name, topic, language, status, education-level, target-audience, etc.)
• Detailed body describing the educational scope
• Updated by the orchestrator as the investigation progresses

Context Files

• See workflows/investigate.md for the investigation workflow
• See workflows/orchestrate.md for agent orchestration
• See ../research/workflows/continue.md for deepening investigations
• See ../research/workflows/status.md for checking status
• See ../research/workflows/archive.md for archival

Guardrails

This skill and all agents it triggers are read-only educational tools.

• The ONLY writable directory is .research/ and .requests/
• No code execution that modifies the system
• No privilege escalation
• No network requests to external APIs on behalf of targets
• No destructive filesystem operations
• Every factual claim requires an identifiable source
• Controversial topics must include multiple perspectives
• Medical/legal/financial topics include appropriate disclaimers