clavix 4.12.0 → 5.0.0

package/dist/templates/slash-commands/_canonical/execute.md

@@ -71,19 +71,18 @@ Me:     [Finds your latest prompt]
 Me:     "Done! Here's what I built..."
 ```
 
-### The Detailed Version
+### The Detailed Version (v5 Agentic-First)
 
 **Step 1: I find your prompt**
 
-I automatically run these commands (you don't need to):
-- `clavix execute --latest` - Get most recent prompt
-- Or `clavix execute --latest --standard` - Get latest standard depth prompt
-- Or `clavix execute --latest --comprehensive` - Get latest comprehensive depth prompt
-- Or `clavix execute --id <id>` - Get specific prompt
+I read directly from the file system:
+- List `.clavix/outputs/prompts/*.md` to find saved prompts
+- Get the most recent one (by timestamp in filename or frontmatter)
+- Read the prompt file: `.clavix/outputs/prompts/<id>.md`
 
 **Step 2: I read and understand**
 
-I'll parse the prompt file, extract:
+I parse the prompt file and extract:
 - The objective (what to build)
 - Requirements (specifics to implement)
 - Technical constraints (how to build it)
@@ -91,7 +90,7 @@ I'll parse the prompt file, extract:
 
 **Step 3: I implement everything**
 
-This is where I actually write code:
+This is where I actually write code using my native tools:
 - Create new files as needed
 - Modify existing files
 - Write functions, components, classes
@@ -99,10 +98,10 @@ This is where I actually write code:
 
 **Step 4: I verify automatically**
 
-After building, I run verification myself:
-- `clavix verify --latest` - Check the implementation
-- Run any automated tests
-- Build/compile to ensure no errors
+After building, I verify by:
+- Running tests (if test suite exists)
+- Building/compiling to ensure no errors
+- Checking requirements from the checklist
 
 **Step 5: I report results**
 
@@ -176,52 +175,43 @@ I'll explain what's wrong and what you might need to do:
 
 ---
 
-## Prompt Management (Commands I Run)
-
-These are commands I execute automatically - you don't need to run them.
+## Prompt Management (v5 Agentic-First)
 
 **Where prompts live:**
-- All prompts: `.clavix/outputs/prompts/`
-
-### Commands I Use (Reference)
-
-| What I Do | Command I Run |
-|-----------|---------------|
-| List saved prompts | `clavix prompts list` |
-| Get latest prompt | `clavix execute --latest` |
-| Get specific prompt | `clavix execute --id <id>` |
-| Run verification | `clavix verify --latest` |
-| Clean up executed | `clavix prompts clear --executed` |
-| Clean up stale (>30d) | `clavix prompts clear --stale` |
+- All prompts: `.clavix/outputs/prompts/*.md`
+- Metadata: In frontmatter of each `.md` file
 
-### Automatic Cleanup
+### How I Access Prompts (Native Tools)
 
-After I finish implementing and verification passes:
-- I clean up executed prompts automatically
-- Old prompts (>30 days) get flagged as stale for removal
-- I keep your storage tidy
+| What I Do | How I Do It |
+|-----------|-------------|
+| List saved prompts | List `.clavix/outputs/prompts/*.md` files |
+| Get latest prompt | Find newest file by timestamp in filename |
+| Get specific prompt | Read `.clavix/outputs/prompts/<id>.md` |
+| Mark as executed | Edit frontmatter: `executed: true` |
+| Clean up executed | Delete files where frontmatter has `executed: true` |
 
 ### The Prompt Lifecycle
 
 ```
-1. YOU CREATE   →  /clavix:improve
-2. I EXECUTE    →  /clavix:execute (you are here)
+1. YOU CREATE   →  /clavix:improve (saves to .clavix/outputs/prompts/<id>.md)
+2. I EXECUTE    →  /clavix:execute (you are here) - I read and implement
 3. I VERIFY     →  Automatic verification
-4. I CLEANUP    →  Remove executed prompts
+4. MARK DONE    →  I update frontmatter with executed: true
 ```
 
 ---
 
 ## Finding Your Way Around
 
-Need to see what projects exist or check progress? I use these commands:
+Need to see what projects exist or check progress? I read the file system:
 
-| What I Need | Command I Run |
+| What I Need | How I Find It |
 |-------------|---------------|
-| See all projects | `clavix list` |
-| Check a specific project | `clavix show --output <project>` |
-| See active sessions | `clavix list --sessions` |
-| Find archived work | `clavix list --archived` |
+| See all projects | List directories in `.clavix/outputs/` |
+| Check a specific project | Read `.clavix/outputs/<project>/` files |
+| See saved prompts | List `.clavix/outputs/prompts/*.md` files |
+| Find archived work | List `.clavix/outputs/archive/` |
 
 ---
 
@@ -271,7 +261,7 @@ If I can't get verification to pass after trying:
 
 ---
 
-## Agent Transparency (v4.11)
+## Agent Transparency (v5.0)
 
 ### CLI Reference (Commands I Execute)
 {{INCLUDE:agent-protocols/cli-reference.md}}

package/dist/templates/slash-commands/_canonical/implement.md

@@ -65,9 +65,9 @@ Me:     "All tasks complete! Your project is built."
 
 ### The Detailed Version
 
-**First time I run:**
+**First time I run (v5 Agentic-First):**
 
-1. **I check for your task list** - Load tasks.md from your PRD folder
+1. **I find your task list** - Read `tasks.md` from your PRD folder (`.clavix/outputs/<project>/tasks.md`)
 2. **I ask about git commits** (only if you have lots of tasks):
    > "You've got 12 tasks. Want me to create git commits as I go?
    >
@@ -78,36 +78,32 @@ Me:     "All tasks complete! Your project is built."
    >
    > Which do you prefer? (I'll default to 'none' if you don't care)"
 
-3. **I initialize tracking** - Run `clavix implement` to set up progress tracking
-4. **I start building** - First incomplete task
+3. **I start building** - First incomplete task
 
 **Each task I work on:**
 
 1. **Read the task** - Understand what needs to be built
 2. **Check the PRD** - Make sure I understand the requirements
-3. **Implement it** - Write code, create files, build features
-4. **Mark it complete** - Run `clavix task-complete {task-id}` automatically
-5. **Move to next** - The command shows me what's next
+3. **Implement it** - Write code, create files, build features using my native tools
+4. **Mark it complete** - Use Edit tool to change `- [ ]` to `- [x]` in tasks.md
+5. **Move to next** - Find the next incomplete task
 
 **If we get interrupted:**
 
 No problem! Just run `/clavix:implement` again and I pick up where we left off.
 The checkboxes in tasks.md track exactly what's done.
 
-## ⚠️ Critical Command: task-complete
+## ⚠️ How I Mark Tasks Complete (v5 Agentic-First)
 
-**After finishing EACH task, I MUST run:**
-```bash
-clavix task-complete <task-id>
-```
+**After finishing EACH task, I use my Edit tool to update tasks.md:**
 
-**Why this matters:**
-- Updates tasks.md automatically (checkboxes)
-- Tracks progress correctly in config
-- Triggers git commits (if enabled)
-- Shows me the next task
+Change: `- [ ] Task description` → `- [x] Task description`
 
-**NEVER manually edit tasks.md checkboxes** - always use this command.
+**Why this matters:**
+- Updates tasks.md directly (checkboxes)
+- Progress is tracked by counting checkboxes
+- Git commits (if enabled) are created with my Bash tool
+- I read tasks.md to find the next incomplete task
 
 ---
 
@@ -115,27 +111,19 @@ clavix task-complete <task-id>
 
 **I handle this automatically - you don't need to do anything.**
 
-### What Happens Behind the Scenes
-
-After I finish implementing a task, I run:
-```bash
-clavix task-complete {task-id}
-```
-
-This does several things:
-- Updates the checkbox in tasks.md ([ ] → [x])
-- Tracks progress in the config file
-- Creates a git commit (if you enabled that)
-- Shows me the next task
+### What Happens (v5 Agentic-First)
 
-### Why I Don't Edit Checkboxes Manually
+After I finish implementing a task:
 
-The command keeps everything in sync. If I edited the file directly, the progress tracking could get confused. Trust the system!
+1. **I use Edit tool** to change `- [ ]` to `- [x]` in tasks.md
+2. **I count progress** by reading tasks.md and counting checkboxes
+3. **I commit** (if you enabled that) using git commands
+4. **I find next task** by scanning for the next `- [ ]` in tasks.md
 
 ### What You'll See
 
 ```
-✓ Task complete: "Set up project structure" (phase-1-setup-1)
+✓ Task complete: "Set up project structure"
 
 Progress: 2/8 tasks (25%)
 
@@ -160,14 +148,14 @@ Starting now...
 
 ## Finding Your Way Around
 
-Need to see what projects exist or check progress? I use these commands:
+Need to see what projects exist or check progress? I read the file system:
 
-| What I Need | Command I Run |
+| What I Need | How I Find It |
 |-------------|---------------|
-| See all projects | `clavix list` |
-| Check a specific project | `clavix show --output <project>` |
-| See active sessions | `clavix list --sessions` |
-| Find archived work | `clavix list --archived` |
+| See all projects | List directories in `.clavix/outputs/` |
+| Check a specific project | Read `.clavix/outputs/<project>/` files |
+| See task progress | Read `.clavix/outputs/<project>/tasks.md` |
+| Find archived work | List `.clavix/outputs/archive/` |
 
 ## When I Can't Continue (Blocked Tasks)
 
@@ -314,7 +302,7 @@ You don't need to remember these - I handle all the tracking.
 
 ---
 
-## Agent Transparency (v4.9)
+## Agent Transparency (v5.0)
 
 ### Workflow State Detection
 {{INCLUDE:agent-protocols/state-awareness.md}}
@@ -348,14 +336,14 @@ You don't need to remember these - I handle all the tracking.
 > - Did you run `/clavix:plan` first?
 > - Is there a PRD folder in .clavix/outputs/?"
 
-### "Task command not working"
+### "Can't update tasks.md"
 
-**What happened:** The `clavix task-complete` command isn't recognized.
+**What happened:** I couldn't edit the tasks.md file to mark tasks complete.
 
 **What I'll do:**
-> "Having trouble with the task command. Let me check your Clavix version..."
+> "Having trouble updating tasks.md. Let me check permissions..."
 >
-> If it's outdated, I'll suggest: "Try `npm install -g clavix@latest` to update"
+> Common fixes: Check file permissions, ensure .clavix/outputs/ is writable
 
 ### "Can't find that task ID"
 

package/dist/templates/slash-commands/_canonical/improve.md

@@ -373,9 +373,8 @@ This is a BLOCKING checkpoint. You cannot proceed to the final message until sav
 |------|--------|-------------|--------------|
 | 1 | Create directory | Write tool (create parent dirs) | Directory exists |
 | 2 | Generate prompt ID | Format: `{std\|comp}-YYYYMMDD-HHMMSS-<random>` | ID is unique |
-| 3 | Write prompt file | **Write tool** | File created |
-| 4 | Update .index.json | **Write tool** | Entry added |
-| 5 | **VERIFY: Read back files** | **Read tool** | Both files readable |
+| 3 | Write prompt file with frontmatter | **Write tool** | File created |
+| 4 | **VERIFY: Read back file** | **Read tool** | File readable |
 
 **⚠️ WARNING:** If you output "saved" without completing verification, you are LYING to the user.
 
@@ -436,63 +435,26 @@ originalPrompt: <user's original prompt text>
 <Insert edge cases>
 ```
 
-### Step 4: Update Index File (Write Tool)
-
-**Use the Write tool** to update the index at `.clavix/outputs/prompts/.index.json`:
-
-**If index file doesn't exist**, create it with:
-```json
-{
-  "version": "2.0",
-  "prompts": []
-}
-```
-
-**Then add a new metadata entry** to the `prompts` array:
-```json
-{
-  "id": "<prompt-id>",
-  "filename": "<prompt-id>.md",
-  "depthUsed": "standard|comprehensive",
-  "timestamp": "<ISO-8601 timestamp>",
-  "createdAt": "<ISO-8601 timestamp>",
-  "path": ".clavix/outputs/prompts/<prompt-id>.md",
-  "originalPrompt": "<user's original prompt text>",
-  "executed": false,
-  "executedAt": null
-}
-```
-
 ---
 
 ## ✅ VERIFICATION (REQUIRED - Must Pass Before Final Output)
 
-**After completing Steps 1-4, you MUST verify the save succeeded.**
+**After completing Steps 1-3, you MUST verify the save succeeded.**
 
-### Verification Step A: Read the Prompt File
+### Verification: Read the Prompt File
 
 Use the **Read tool** to read the file you just created:
 - Path: `.clavix/outputs/prompts/<your-prompt-id>.md`
 
-**If Read fails:** ⛔ STOP - Saving failed. Retry Steps 3-4.
-
-### Verification Step B: Read the Index File
-
-Use the **Read tool** to read the index:
-- Path: `.clavix/outputs/prompts/.index.json`
-
-**Confirm:** Your prompt ID appears in the `prompts` array.
-
-**If not found:** ⛔ STOP - Index update failed. Retry Step 4.
+**If Read fails:** ⛔ STOP - Saving failed. Retry Step 3.
 
 ### Verification Checklist
 
 Before outputting final message, confirm ALL of these:
 
 - [ ] I used the **Write tool** to create `.clavix/outputs/prompts/<id>.md`
-- [ ] I used the **Write tool** to update `.clavix/outputs/prompts/.index.json`
 - [ ] I used the **Read tool** to verify the prompt file exists and has content
-- [ ] I used the **Read tool** to verify my prompt ID is in .index.json
+- [ ] The file has valid frontmatter with id, timestamp, and executed: false
 - [ ] I know the **exact file path** I created (not a placeholder)
 
 **If ANY checkbox is unchecked: ⛔ STOP and complete the missing step.**
@@ -538,13 +500,13 @@ Wait for the user to decide what to do next.
 - `/clavix:start` - Conversational exploration before prompting
 - `/clavix:verify` - Verify implementation against checklist
 
-**CLI commands:**
-- `clavix prompts list` - View saved prompts
-- `clavix prompts clear --executed` - Clean up executed prompts
+**Managing saved prompts:**
+- List prompts: `ls .clavix/outputs/prompts/*.md`
+- Prompt files: `.clavix/outputs/prompts/<id>.md` (metadata in frontmatter)
 
 ---
 
-## Agent Transparency (v4.11)
+## Agent Transparency (v5.0)
 
 ### CLI Reference (Commands I Execute)
 {{INCLUDE:agent-protocols/cli-reference.md}}
@@ -586,10 +548,9 @@ Wait for the user to decide what to do next.
 mkdir -p .clavix/outputs/prompts
 ```
 
-**Error: Index file corrupted or invalid JSON**
-```bash
-echo '{"version":"2.0","prompts":[]}' > .clavix/outputs/prompts/.index.json
-```
+**Error: Prompt file has invalid frontmatter**
+- Re-save the prompt file with valid YAML frontmatter
+- Ensure id, timestamp, and executed fields are present
 
 ### Issue: Wrong depth auto-selected
 **Cause**: Borderline quality score