homunculus-code 0.3.3 → 0.3.4

package/commands/hm-setup.md

@@ -8,19 +8,33 @@ Guide the user through defining their goals and generate `architecture.yaml`.
 
 You are helping the user set up Homunculus — a self-evolving AI assistant. Your job is to understand their project and goals, then generate a goal tree.
 
-### Step 1: Understand the Project
+**CRITICAL: Ask ONE question at a time. Wait for the answer before asking the next. Never batch multiple questions in one message.**
 
-Ask the user (conversationally, not a form):
-- What is this project? (e.g., SaaS app, CLI tool, personal project)
-- What do they spend most time on? (e.g., debugging, writing tests, deploying)
-- What frustrates them? (e.g., regressions, slow CI, repetitive tasks)
-- What would they improve if they had infinite time?
+### Step 1: Ask about the project (ONE question)
 
-Keep it natural — 2-3 questions max, adapt based on answers.
+Start with:
+> "What kind of project is this? (e.g., web app, CLI tool, API, personal project)"
 
-### Step 2: Propose Goals
+Wait for answer.
 
-Based on their answers, propose 3-5 goals with sub-goals. Present them clearly:
+### Step 2: Ask about pain points (ONE question)
+
+Based on their answer, ask:
+> "What do you spend the most time on that you wish was better? (e.g., debugging, testing, deployment, keeping up with updates)"
+
+Wait for answer.
+
+### Step 3: Ask about priorities (ONE question)
+
+Based on their answer, ask ONE more targeted question. Examples:
+- "If your AI assistant could fix one thing overnight, what would it be?"
+- "What breaks most often in your workflow?"
+
+Adapt this question based on what they already told you. Wait for answer.
+
+### Step 4: Propose goals
+
+Based on ALL their answers, propose 3-5 goals with sub-goals:
 
 ```
 Based on what you told me, here's your goal tree:
@@ -30,25 +44,26 @@ Based on what you told me, here's your goal tree:
 │   ├── testing — Every change has tests
 │   └── review — Catch issues before merge
 ├── productivity — Move faster
-│   ├── automation — Automate repetitive work
 │   └── debugging — Find root causes faster
-└── knowledge — Stay up to date
+└── knowledge — Stay current
     └── tool_updates — Track useful updates
+
+Does this look right? Want to add, remove, or change anything?
 ```
 
-Ask: "Does this look right? Want to add, remove, or change anything?"
+Wait for confirmation or changes.
 
-### Step 3: Generate architecture.yaml
+### Step 5: Generate architecture.yaml
 
 Once confirmed, generate `architecture.yaml` with:
 - `purpose` for every goal
 - `metrics` where measurable (use reasonable defaults)
-- `health_check` where possible (shell commands that exit 0 = healthy)
-- `realized_by: # will evolve` for all implementations (the system will fill these in)
+- `health_check` where possible (use the project's actual tech stack: `npm test`, `pytest`, `go test`, etc.)
+- `realized_by: # will evolve` for all implementations
 
 Write the file using the Write tool.
 
-### Step 4: Confirm
+### Step 6: Confirm
 
 ```
 ✅ architecture.yaml created with N goals!
@@ -58,10 +73,10 @@ patterns will be auto-extracted. Run /hm-night anytime to
 trigger an evolution cycle.
 ```
 
-## Important
+## Rules
 
-- Keep the conversation SHORT (under 5 back-and-forth)
+- **ONE question per message. Never ask two questions at once.**
+- Keep the whole setup under 5 back-and-forth messages
 - Generate PRACTICAL goals, not abstract ones
-- Use the project's actual tech stack for health checks (e.g., `npm test`, `pytest`, `go test`)
-- Don't overwhelm — 3-5 top-level goals is ideal for a start
-- Goals can always be refined later
+- Don't overwhelm — 3-5 top-level goals is ideal
+- Goals can always be refined later with `/hm-setup` again

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "homunculus-code",
-  "version": "0.3.3",
+  "version": "0.3.4",
   "description": "A self-evolving AI assistant that grows smarter every night",
   "bin": {
     "homunculus-code": "./bin/cli.js"