agestra 4.7.0 → 4.8.0

package/skills/idea.md

@@ -1,212 +1,307 @@
----
-name: agestra-idea
-description: >
-  Use when discovering improvements, comparing with similar projects, collecting user feedback,
-  exploring new features, researching what to build, or validating ideas. Triggers on:
-  "find improvements", "what should I add", "compare with competitors", "what are users asking for",
-  "explore ideas", "feature ideas", "what's missing", "is this worth building", "what do users want",
-  "what problem does this solve", "who would use this", "what should I focus on next",
-  "개선점", "뭐 추가하면 좋을까", "아이디어", "유사 프로젝트", "뭐가 부족해",
-  "이거 만들 가치가 있어?", "다음에 뭘 해야 할까", "비슷한 도구",
-  "改善", "アイデア", "改进", "想法"
----
-
-## Purpose
-
-Idea and improvement discovery. Research similar projects, collect user complaints and feature requests, compare capabilities, and generate actionable suggestions.
-
-## Scope
-
-**Mode A: Existing project** — The codebase has a README or meaningful code.
-Research improvements, missing features, and competitive gaps for this project.
-
-**Mode B: New project** — The codebase is empty/new, but the user has a seed idea (e.g., "I want to build a writing tool").
-Research the landscape: what already exists, what users complain about, what gaps remain.
-
-**Out of scope:** Requests with no seed idea at all (e.g., "what should I build?"). You need at least a domain or concept to anchor research. Ask for one:
-
-> "I need at least a rough idea — a domain, a tool type, or a problem you want to solve. For example: 'a writing tool', 'a CLI for deployment', 'something for managing bookmarks'."
-
-## Workflow
-
-### Phase 1: Clarity Gate
-
-Before researching, understand what the user needs through targeted questions. Ask ONE question at a time. Communicate in the user's language.
-
-**Step 1: Determine mode.**
-- If the codebase has a README or meaningful code → Mode A (existing project)
-- If the codebase is empty/new but user has a seed idea → Mode B (new project)
-
-**Step 2: Mode-specific interview.**
-
-**Mode A — Existing project:**
-
-| Dimension | Question | Purpose |
-|-----------|----------|---------|
-| Direction | "What aspect are you looking to improve? (features, UX, performance, integrations, DX)" | Narrow the research scope |
-| Audience | "Who are your current users? What do they use it for most?" | Target the right competitors |
-| Feedback | "Have you received any complaints or feature requests?" | Direct pain point input |
-| Competition | "Are there specific competitors or similar tools you're aware of?" | Seed the research |
-| Strength | "What do you consider your project's unique strength?" | Avoid suggesting what already works |
-| Constraints | "Any areas you don't want to change or can't change?" | Set research boundaries |
-
-After gathering context:
-- Read the project's README and key files to understand what it does
-- Use Glob and Grep to map the current feature set
-- Identify the project's category and target audience
-
-**Mode B — New project:**
-
-| Dimension | Question | Purpose |
-|-----------|----------|---------|
-| Problem | "What problem are you trying to solve?" | Core motivation |
-| Audience | "Who would use this? What's the target audience?" | Market focus |
-| Form | "How do you envision it? (CLI, web app, library, service, plugin)" | Shape the research |
-| Inspiration | "What inspired this? Have you seen something similar?" | Seed the research |
-| Core | "What's the single most important thing it must do well?" | Prioritization anchor |
-| Boundary | "What should it NOT be? Where do you draw the line?" | Scope limits |
-
-**Early exit:** If the user provides enough context upfront (specific competitors, clear scope, concrete goals), skip remaining questions and proceed to Phase 2. Do not force unnecessary rounds.
-
-### Phase 2: Check environment and select mode
-
-Call `environment_check` to determine available providers.
-
-Present mode selection (in the user's language):
-
-| Option | Condition | Description |
-|--------|-----------|-------------|
-| **Claude only** | Always | 플러그인 전문 에이전트가 단독으로 개선점 탐색 |
-| **각자 독립** | 1+ provider available | 에이전트 팀(moderator + ideator + 외부 AI)이 각자 독립 탐색 후 취합 문서 작성 |
-| **끝장토론** | 1+ provider available | 각자 독립 탐색 후 취합 문서를 돌아가며 분석/피드백, 모두 동의할 때까지 |
-
-Only show options whose conditions are met. If no providers are available, run Claude only.
-
-### Phase 3: Execute
-
-#### If "Claude only":
-Proceed to Phase 4 (Research) directly using `agestra-ideator` agent.
-
-#### If "각자 독립":
-
-**팀 구성:** `agestra-moderator` (리더) + `agestra-ideator` (Claude) + 사용 가능한 외부 AI (gemini, codex, ollama 등)
-
-1. In parallel:
-   - Spawn the `agestra-ideator` agent for Claude's independent improvement research.
-     After the agent completes, save Claude's result as a document via `workspace_create_document`:
-     - **title:** `Idea Exploration — claude/ideator`
-     - **metadata:** `{ "Provider": "claude/ideator", "Task": "{topic}", "Mode": "Independent" }`
-     - **content:** The ideator agent's full output.
-   - For each available provider, call `ai_chat` with `save_as_document`:
-     - **save_as_document.title:** `Idea Exploration — {provider}`
-     - **save_as_document.metadata:** `{ "Task": "{topic}", "Mode": "Independent" }`
-     - **prompt:** Research improvements for [topic]. Look at similar projects, common user complaints, missing features, and opportunities. For each suggestion, provide: title, category (UX/Performance/Feature/Integration/DX), source of the idea, priority (HIGH/MEDIUM/LOW), and a brief description.
-
-2. Collect all document IDs.
-
-3. Spawn the `agestra-moderator` agent in **Independent Aggregation** mode:
-   - Pass the document ID list.
-   - Moderator reads each document, classifies consensus/unique/disputed suggestions.
-   - Moderator creates an aggregated document via `workspace_create_document`.
-
-4. Present summary report to the user:
-   - Key consensus suggestions (1-3 lines).
-   - Notable unique ideas (if any).
-   - Priority disagreements (if any).
-   - Individual + aggregated document IDs.
-
-#### If "끝장토론":
-
-**팀 구성:** `agestra-moderator` (리더) + `agestra-ideator` (Claude) + 사용 가능한 외부 AI (gemini, codex, ollama 등)
-
-1. Execute "각자 독립" steps 1-3 above (independent work + initial aggregation).
-   - The moderator's integrated document becomes the starting document.
-
-2. Document review rounds (no max — until all agree):
-   a. Moderator sends the current document to each AI for review:
-      - Claude: spawn `agestra-ideator` → analyze document → section-by-section feedback
-      - Other providers: `agent_debate_turn` with the document as prompt, requesting agree/disagree per section
-   b. Moderator collects all feedback.
-   c. Classify: agree/disagree per section per provider.
-   d. Revise document incorporating disagreement feedback.
-   e. If all providers agree on all sections → consensus reached.
-   f. If not → next round with revised document.
-   g. **Every 10 rounds:** Ask the user whether to continue or stop with current state.
-
-3. Present the final document:
-   - Consensus sections: marked as agreed
-   - Disputed sections: show split positions with each provider's rationale
-
-### Phase 4: Research (Claude only mode)
-
-If running Claude only, execute the following research phases:
-
-#### 4a: Research Similar Projects
-- Use WebSearch to find similar tools, libraries, and projects
-- Look for: direct competitors, adjacent tools, inspirational projects
-- Collect names, URLs, and key differentiators
-
-#### 4b: Collect Pain Points
-- WebSearch for complaints about similar tools (GitHub issues, forums, discussions)
-- WebFetch relevant issue pages and discussion threads
-- Identify recurring themes in user feedback
-- Note what users wish existed but doesn't
-
-#### 4c: Feature Comparison
-Build a comparison table:
-
-| Feature | This Project | Competitor A | Competitor B |
-|---------|-------------|-------------|-------------|
-| Feature 1 | Yes/No | Yes/No | Yes/No |
-
-#### 4d: Generate Suggestions
-For each suggestion:
-- **Title** — clear, actionable name
-- **Category** — UX, Performance, Feature, Integration, DX
-- **Source** — where this idea came from (competitor, user complaint, own analysis)
-- **Priority** — HIGH / MEDIUM / LOW with rationale
-- **Effort** — estimated complexity
-- **Description** — what it does and why it matters
-
-#### 4e: Prioritized Recommendations
-Present a ranked list:
-1. **Quick wins** — high impact, low effort
-2. **Strategic investments** — high impact, high effort
-3. **Nice-to-haves** — low impact, low effort
-
-## Output Format
-
-```markdown
-## Research Summary
-
-### Similar Projects
-(list with URLs and key features)
-
-### User Pain Points
-(categorized complaints from research)
-
-### Feature Comparison
-(table)
-
-### Recommendations
-
-#### Quick Wins
-1. ...
-
-#### Strategic Investments
-1. ...
-
-#### Nice-to-Haves
-1. ...
-
-### Sources
-- [Source 1](url)
-- [Source 2](url)
-```
-
-## Constraints
-
-- Always include source URLs for claims about other projects.
-- Do not fabricate features of competitors — verify via web research.
-- Prioritize actionable suggestions over theoretical improvements.
-- Communicate in the user's language.
+---
+name: agestra-idea
+description: >
+  Use when discovering improvements, comparing with similar projects, collecting user feedback,
+  exploring new features, researching what to build, or validating ideas. Triggers on:
+  "find improvements", "what should I add", "compare with competitors", "what are users asking for",
+  "explore ideas", "feature ideas", "what's missing", "is this worth building", "what do users want",
+  "what problem does this solve", "who would use this", "what should I focus on next",
+  "개선점", "뭐 추가하면 좋을까", "아이디어", "유사 프로젝트", "뭐가 부족해",
+  "이거 만들 가치가 있어?", "다음에 뭘 해야 할까", "비슷한 도구",
+  "改善", "アイデア", "改进", "想法"
+---
+
+## Purpose
+
+Idea and improvement discovery. Research similar projects, collect user complaints and feature requests, compare capabilities, and generate actionable suggestions.
+
+## Scope
+
+**Mode A: Existing project** — The codebase has a README or meaningful code.
+Research improvements, missing features, and competitive gaps for this project.
+
+**Mode B: New project** — The codebase is empty/new, but the user has a seed idea (e.g., "I want to build a writing tool").
+Research the landscape: what already exists, what users complain about, what gaps remain. Define what to build and why it's worth building.
+
+**Out of scope:** Requests with no seed idea at all (e.g., "what should I build?"). You need at least a domain or concept to anchor research. Ask for one:
+
+> "I need at least a rough idea — a domain, a tool type, or a problem you want to solve. For example: 'a writing tool', 'a CLI for deployment', 'something for managing bookmarks'."
+
+## Workflow
+
+### Phase 1: Clarity Gate
+
+Before researching, understand what the user needs through targeted questions. Ask ONE question at a time. Communicate in the user's language.
+
+**Step 1: Determine mode.**
+- If the codebase has a README or meaningful code → Mode A (existing project)
+- If the codebase is empty/new but user has a seed idea → Mode B (new project)
+
+**Step 2: Mode-specific interview.**
+
+**Mode A — Existing project:**
+
+| Dimension | Question | Purpose |
+|-----------|----------|---------|
+| Direction | "What aspect are you looking to improve? (features, UX, performance, integrations, DX)" | Narrow the research scope |
+| Audience | "Who are your current users? What do they use it for most?" | Target the right competitors |
+| Feedback | "Have you received any complaints or feature requests?" | Direct pain point input |
+| Competition | "Are there specific competitors or similar tools you're aware of?" | Seed the research |
+| Strength | "What do you consider your project's unique strength?" | Avoid suggesting what already works |
+| Constraints | "Any areas you don't want to change or can't change?" | Set research boundaries |
+
+After gathering context:
+- Read the project's README and key files to understand what it does
+- Use Glob and Grep to map the current feature set
+- Identify the project's category and target audience
+
+**Mode B — New project:**
+
+| Dimension | Question | Purpose |
+|-----------|----------|---------|
+| Problem | "What problem are you trying to solve? What frustration triggered this idea?" | Core motivation — this anchors all research |
+| Audience | "Who would use this? Describe the ideal user." | Determines which competitors and pain points to research |
+| Form | "How do you envision it? (CLI, web app, mobile, library, service, plugin, desktop)" | Shape the research — different forms have different competitive landscapes |
+| Inspiration | "What inspired this? Have you seen something similar that's close but not quite right?" | Seeds competitor research with concrete starting points |
+| Core | "If it could only do ONE thing perfectly, what would that be?" | Prioritization anchor — prevents scope creep in research |
+| Boundary | "What should it NOT be? What's explicitly out of scope?" | Prevents researching irrelevant directions |
+| Differentiator | "Why would someone use yours instead of what already exists?" | Tests whether the user has a hypothesis to validate |
+
+**Early exit:** If the user provides enough context upfront (specific competitors, clear scope, concrete goals), skip remaining questions and proceed to Phase 2. Do not force unnecessary rounds.
+
+### Phase 2: Check environment and select mode
+
+Call `environment_check` to determine available providers.
+
+- If **no providers are available**: proceed to Phase 3 (Research) directly using `agestra-ideator` agent (Claude only).
+- If **1+ providers are available**: proceed to 끝장토론 execution below.
+
+### Phase 3: Execute 끝장토론
+
+**팀 구성:** `agestra-moderator` (리더) + `agestra-ideator` (Claude) + 사용 가능한 외부 AI (gemini, codex 등)
+
+The `ai_chat` prompt for external providers differs by mode. Use the interview answers collected in Phase 1 to construct a context-rich prompt.
+
+1. Independent work + initial aggregation:
+
+   a. In parallel:
+      - Spawn the `agestra-ideator` agent for Claude's independent research.
+        After the agent completes, save Claude's result as a document via `workspace_create_document`:
+        - **title:** `Idea Exploration — claude/ideator`
+        - **metadata:** `{ "Provider": "claude/ideator", "Task": "{topic}", "Mode": "Independent" }`
+        - **content:** The ideator agent's full output.
+      - For each available provider, call `ai_chat` with `save_as_document`:
+        - **save_as_document.title:** `Idea Exploration — {provider}`
+        - **save_as_document.metadata:** `{ "Task": "{topic}", "Mode": "Independent" }`
+        - **prompt:** Use the Mode A or Mode B prompt below.
+
+   **Mode A prompt for external providers:**
+
+   > Context from user interview:
+   > - Direction: {user's answer}
+   > - Audience: {user's answer}
+   > - Known competitors: {user's answer}
+   > - Project strengths: {user's answer}
+   > - Constraints: {user's answer}
+   >
+   > Analyze this existing project and generate exactly 5 improvement ideas.
+   > For each idea, provide this exact structure:
+   >
+   > 1. **Title** — clear, actionable name
+   > 2. **Category** — one of: Feature, UX, Performance, Integration, DX, Security, Documentation
+   > 3. **Evidence** — specific file, pattern, or code area that reveals this opportunity
+   > 4. **Description** — what to change and why it matters
+   > 5. **Effort** — trivial / small / medium / large / complex, with brief justification
+   > 6. **Priority** — HIGH / MEDIUM / LOW with rationale
+   > 7. **Implementation approach** — concrete first steps
+   >
+   > Rules:
+   > - Every idea must be grounded in actual code or project structure (no generic advice)
+   > - Include a mix of effort levels (not all "large")
+   > - Do not suggest changes to areas the user marked as constraints
+
+   **Mode B prompt for external providers:**
+
+   > Context from user interview:
+   > - Problem to solve: {user's answer}
+   > - Target audience: {user's answer}
+   > - Envisioned form: {user's answer}
+   > - Inspiration/similar tools: {user's answer}
+   > - Core must-have: {user's answer}
+   > - Out of scope: {user's answer}
+   > - Differentiator hypothesis: {user's answer}
+   >
+   > Research the landscape for this new project idea and provide:
+   >
+   > **Part 1: Competitive Landscape (3-5 existing tools)**
+   > For each competitor:
+   > - Name and URL
+   > - What it does well
+   > - What users complain about (cite sources if possible)
+   > - How the user's idea differs
+   >
+   > **Part 2: Opportunity Analysis**
+   > - Gaps in existing tools that no one fills
+   > - Most common user frustrations across competitors
+   > - Underserved audience segments
+   >
+   > **Part 3: Feature Recommendations (5-7 ideas)**
+   > For each feature:
+   > 1. **Title** — clear name
+   > 2. **Category** — Core (must-have for v1) / Differentiator (sets it apart) / Growth (future expansion)
+   > 3. **Why** — how this addresses a real gap or pain point found in research
+   > 4. **Effort** — trivial / small / medium / large / complex
+   > 5. **Priority** — P0 (v1 essential) / P1 (v1 nice-to-have) / P2 (post-launch)
+   > 6. **Risk** — what could go wrong or make this not worth building
+   >
+   > **Part 4: Viability Assessment**
+   > - Is the user's differentiator hypothesis valid? Why or why not?
+   > - What's the biggest risk for this project?
+   > - Recommended MVP scope (3-5 features max)
+   >
+   > Be thorough and specific. Generic advice like "make it user-friendly" is not useful.
+
+   b. Collect all document IDs.
+
+   c. Spawn the `agestra-moderator` agent in **Independent Aggregation** mode:
+      - Pass the document ID list.
+      - Moderator reads each document, classifies consensus/unique/disputed suggestions.
+      - Moderator creates an aggregated document. This becomes the starting document.
+
+2. Start a turn-based debate session with `agent_debate_create`.
+
+3. Document review rounds (no max — until all agree):
+   a. Moderator sends the current document to each AI for review.
+   b. Moderator collects all feedback and updates the working document.
+   c. If all providers agree on all sections → consensus reached.
+   d. If not → next round with the revised document.
+   e. **Every 10 rounds:** Ask the user whether to continue or stop with current state.
+
+4. When consensus is reached, record the final synthesis with `agent_debate_conclude`.
+
+5. Present the final document:
+   - Consensus sections: marked as agreed
+   - Disputed sections: show split positions with each provider's rationale
+
+### Phase 4: Research (Claude only mode)
+
+If running Claude only, execute the following research phases:
+
+#### Mode A: Existing project research
+
+**4a: Codebase Analysis**
+- Read project README, key files, and directory structure
+- Use Glob and Grep to map current features and patterns
+- Identify code smells, gaps, and improvement areas grounded in actual code
+
+**4b: Research Similar Projects**
+- Use WebSearch to find similar tools, libraries, and projects
+- Focus on competitors the user mentioned in the interview
+- Collect names, URLs, and key differentiators
+
+**4c: Collect Pain Points**
+- WebSearch for complaints about similar tools (GitHub issues, forums, discussions)
+- WebFetch relevant issue pages and discussion threads
+- Identify recurring themes in user feedback
+
+**4d: Generate 5 structured ideas**
+Follow the Mode A prompt structure above (Title, Category, Evidence, Description, Effort, Priority, Implementation approach).
+
+**4e: Prioritized Recommendations**
+Present a ranked list:
+1. **Quick wins** — high impact, low effort
+2. **Strategic investments** — high impact, high effort
+3. **Nice-to-haves** — low impact, low effort
+
+#### Mode B: New project research
+
+**4a: Competitive Landscape**
+- WebSearch for existing tools in the domain the user described
+- For each competitor: features, pricing, user reviews, GitHub stars/activity
+- WebFetch key competitor pages for detailed feature analysis
+- Cross-reference with the user's inspiration and known similar tools
+
+**4b: User Pain Points**
+- WebSearch for complaints: "{domain} tool frustrations", "{competitor} alternatives", "{problem} solution"
+- Look at GitHub issues, Reddit threads, forum posts, product review sites
+- Categorize by frequency and severity
+
+**4c: Gap Analysis**
+- What do users want that no existing tool provides?
+- Which audience segments are underserved?
+- What technical approaches haven't been tried?
+
+**4d: Generate 5-7 structured feature ideas**
+Follow the Mode B prompt structure above (Title, Category, Why, Effort, Priority, Risk).
+
+**4e: Viability Assessment**
+- Validate or challenge the user's differentiator hypothesis
+- Identify the biggest risk
+- Recommend MVP scope (3-5 features max)
+- Suggest a "litmus test" — one thing the user could build in a weekend to validate demand
+
+## Output Format
+
+### Mode A output:
+
+```markdown
+## Improvement Ideas for {project}
+
+### Quick Wins
+1. **{Title}** [{Category}] — Effort: {effort}
+   {Description}. Evidence: {file/pattern}.
+
+### Strategic Investments
+1. ...
+
+### Nice-to-Haves
+1. ...
+
+### Competitive Comparison
+| Feature | This Project | {Competitor A} | {Competitor B} |
+|---------|-------------|----------------|----------------|
+
+### Sources
+- [Source 1](url)
+```
+
+### Mode B output:
+
+```markdown
+## Research: {project idea}
+
+### Competitive Landscape
+| Tool | Strengths | Weaknesses | How we differ |
+|------|-----------|------------|---------------|
+
+### User Pain Points
+(categorized by frequency)
+
+### Recommended Features
+#### Core (v1 essential)
+1. **{Title}** — {Why}. Effort: {effort}. Risk: {risk}.
+
+#### Differentiators
+1. ...
+
+#### Growth (post-launch)
+1. ...
+
+### Viability Assessment
+- Differentiator valid? {yes/no + reasoning}
+- Biggest risk: {description}
+- Recommended MVP: {3-5 features}
+- Quick validation: {litmus test suggestion}
+
+### Sources
+- [Source 1](url)
+```
+
+## Constraints
+
+- Always include source URLs for claims about other projects.
+- Do not fabricate features of competitors — verify via web research.
+- Every Mode A idea must cite specific code evidence (file, pattern, or area).
+- Mode B ideas must trace back to research findings (competitor gap, user complaint, or market opportunity).
+- Prioritize actionable suggestions over theoretical improvements.
+- Communicate in the user's language.

package/skills/provider-guide.md

@@ -50,13 +50,14 @@ Full-capability agents. Use for:
 
 ### Text Work (리뷰/설계/아이디어)
 
-Three modes available via `/agestra review`, `/agestra design`, `/agestra idea`:
+Available via `/agestra review`, `/agestra design`, `/agestra idea`:
 
 | Mode | Description | When to Use |
 |------|-------------|-------------|
-| **Claude only** | Specialist agent works alone | Quick analysis, no external AI needed |
-| **각자 독립** | Each AI works independently → moderator aggregates | Want multiple perspectives, fast |
-| **끝장토론** | Independent work + document review rounds until consensus | Need thorough, agreed-upon analysis |
+| **Claude only** | Specialist agent works alone | No external providers available |
+| **끝장토론** | Independent work → aggregation → review rounds until consensus | Providers available (default) |
+
+When providers are enabled, commands go directly to 끝장토론 mode. No mode selection needed.
 
 ### Implementation Work (실제 구현)
 
@@ -104,8 +105,6 @@ Match by **semantic intent**, not literal keywords. These triggers apply in any
 | Second opinion, other perspectives | `ai_compare` or `/agestra review` (각자 독립) | User wants multiple viewpoints on a decision |
 | Validation, verification, cross-check | `agent_cross_validate` | User wants to confirm correctness of work output |
 | Speed up, parallelize, split work | Team API or CLI workers | User wants faster execution or has independent tasks |
-| Past experience, history, previous attempts | `memory_search` or `memory_dead_ends` | User asks about prior work or known issues |
-| Remember this, save for later | `memory_store` | User wants to persist knowledge across sessions |
 | Mention a provider by name (Gemini, Codex, Ollama) | `ai_chat` or `cli_worker_spawn` | Route directly to the named provider |
 | Architecture review, design discussion | `/agestra design` | Structured multi-AI architecture exploration |
 | Compare options, which is better | `ai_compare` | Side-by-side comparison from multiple providers |
@@ -130,21 +129,17 @@ Match by **semantic intent**, not literal keywords. These triggers apply in any
 | `cancel` | Gracefully stop running operations (including CLI workers) with state cleanup |
 | `worker-manage` | List, check, collect, and stop CLI workers |
 
-When "각자 독립" is selected, each AI works independently and `agestra-moderator` aggregates results.
-When "끝장토론" is selected, `agestra-moderator` facilitates document review rounds after independent aggregation.
-
-Commands and hook-triggered suggestions share the same 3-choice pattern (Claude only / 각자 독립 / 끝장토론). Commands are explicit entry points; hooks detect intent from natural language.
+In 끝장토론 mode, each AI works independently first, then `agestra-moderator` aggregates results and facilitates review rounds until consensus.
 
-### Hook-Triggered Choice
+Commands and hook-triggered suggestions go directly to 끝장토론 when providers are available. Commands are explicit entry points; hooks detect intent from natural language.
 
-When an `AGESTRA_SUGGESTION` marker appears from the UserPromptSubmit hook, present these choices:
+### Hook-Triggered Flow
 
-1. **Claude only** — Claude Code handles it alone
-2. **각자 독립** — Each AI works independently, moderator aggregates
-3. **끝장토론** — Independent work + document review rounds until consensus
-4. **Other** — User specifies the approach
+When the UserPromptSubmit hook injects context, proceed directly to **끝장토론** (debate) mode
+using the enabled providers from `providers.config.json`.
 
-Present choices in the user's language. If no providers are available, skip and proceed directly.
+If no config exists, suggest `/agestra setup` first.
+If no providers are enabled, suggest `/agestra setup` to select AIs.
 
 ## Error Handling — 429 Rate Limit
 
@@ -155,11 +150,10 @@ Present choices in the user's language. If no providers are available, skip and
 
 Do NOT wait for rate limit reset.
 
-## Memory System
+## Setup
 
-- Failed approaches are automatically recorded as `dead_end` nodes.
-- Call `memory_dead_ends` before starting work to avoid repeating failed strategies.
-- Call `memory_store` to save findings for future sessions.
+Run `/agestra setup` to select which AIs to use. Selected AIs are saved in `providers.config.json`.
+All providers default to disabled until explicitly enabled via setup.
 
 ## Orchestration Pipeline