prizmkit 1.0.126 → 1.0.128

package/bin/create-prizmkit.js

@@ -36,7 +36,7 @@ program
   .option('--no-pipeline', 'Disable dev-pipeline')
   .option('--rules <preset>', 'Rules preset: recommended, minimal, or none', 'recommended')
   .option('--ai-cli <command>', 'AI CLI executable command (e.g. cbc, claude, claude-internal)')
-  .option('--external-skills <names>', 'Comma-separated external skill names to install (e.g. find-skill,uiuxpromax)')
+  .option('--external-skills <names>', 'Comma-separated external skill names to install (e.g. find-skills,impeccable)')
   .option('--non-interactive', 'Skip interactive prompts (requires --platform)')
   .option('--dry-run', 'Show what would be created without making changes')
   .action(async (directory = '.', options) => {

package/bundled/VERSION.json

@@ -1,5 +1,5 @@
 {
-  "frameworkVersion": "1.0.126",
-  "bundledAt": "2026-03-28T02:42:16.988Z",
-  "bundledFrom": "78687d9"
+  "frameworkVersion": "1.0.128",
+  "bundledAt": "2026-03-28T06:01:48.243Z",
+  "bundledFrom": "ece8a06"
 }

package/bundled/dev-pipeline/lib/common.sh

@@ -43,6 +43,10 @@ try:
     with open('.prizmkit/config.json') as f:
         d = json.load(f)
     v = d.get('ai_cli', '')
+    if not v:
+        p = d.get('platform', '')
+        if p == 'claude': v = 'claude'
+        elif p == 'codebuddy': v = 'cbc'
     if v: print(v)
 except: pass
 " 2>/dev/null || true)
@@ -50,20 +54,20 @@ except: pass
             _raw_cli="$_config_ai_cli"
         elif [[ -n "${CODEBUDDY_CLI:-}" ]]; then
             _raw_cli="$CODEBUDDY_CLI"
-        elif command -v cbc &>/dev/null; then
-            _raw_cli="cbc"
         elif command -v claude &>/dev/null; then
             _raw_cli="claude"
+        elif command -v cbc &>/dev/null; then
+            _raw_cli="cbc"
         else
             echo "ERROR: No AI CLI found. Install CodeBuddy (cbc) or Claude Code (claude)." >&2
             exit 1
         fi
     elif [[ -n "${CODEBUDDY_CLI:-}" ]]; then
         _raw_cli="$CODEBUDDY_CLI"
-    elif command -v cbc &>/dev/null; then
-        _raw_cli="cbc"
     elif command -v claude &>/dev/null; then
         _raw_cli="claude"
+    elif command -v cbc &>/dev/null; then
+        _raw_cli="cbc"
     else
         echo "ERROR: No AI CLI found. Install CodeBuddy (cbc) or Claude Code (claude)." >&2
         exit 1
@@ -84,6 +88,27 @@ except: pass
     export PRIZMKIT_PLATFORM="$PLATFORM"
 }
 
+# prizm_detect_subagents <session_log>
+#
+# Scan session log for subagent spawns and log the result.
+# Uses grep -q for early exit on first match.
+# Requires: USE_STREAM_JSON (from detect_stream_json_support)
+prizm_detect_subagents() {
+    local session_log="$1"
+    [[ -f "$session_log" ]] || return 0
+
+    local has_subagent=false
+    if [[ "$USE_STREAM_JSON" == "true" ]]; then
+        grep -q '"name"[[:space:]]*:[[:space:]]*"Agent"' "$session_log" 2>/dev/null && has_subagent=true
+    else
+        grep -qE '(Tool: Agent|"tool":\s*"Agent"|tool_use.*Agent|subagent_type)' "$session_log" 2>/dev/null && has_subagent=true
+    fi
+
+    if [[ "$has_subagent" == true ]]; then
+        log_info "Subagent calls detected in session"
+    fi
+}
+
 # Common dependency check (jq + python3 + optional CLI in PATH)
 # Args:
 #   $1 - cli command (optional)

package/bundled/dev-pipeline/run-bugfix.sh

@@ -215,7 +215,7 @@ spawn_and_wait_session() {
             if [[ -n "$dirty_files" ]]; then
                 log_info "Auto-committing remaining session artifacts..."
                 git -C "$project_root" add -A 2>/dev/null || true
-                git -C "$project_root" commit --no-verify --amend --no-edit -a 2>/dev/null \
+                git -C "$project_root" commit --no-verify --amend --no-edit 2>/dev/null \
                     || git -C "$project_root" commit --no-verify -m "chore($bug_id): include remaining session artifacts" 2>/dev/null \
                     || true
             fi
@@ -224,6 +224,9 @@ spawn_and_wait_session() {
 
     log_info "Session result: $session_status"
 
+    # Subagent detection
+    prizm_detect_subagents "$session_log"
+
     # Update bug status
     python3 "$SCRIPTS_DIR/update-bug-status.py" \
         --bug-list "$bug_list" \

package/bundled/dev-pipeline/run.sh

@@ -216,7 +216,7 @@ spawn_and_wait_session() {
             uncommitted=$(git -C "$project_root" status --porcelain 2>/dev/null | head -1 || true)
             if [[ -n "$uncommitted" ]]; then
                 log_warn "Session exited cleanly but produced no commits (uncommitted changes found) — auto-committing..."
-                git -C "$project_root" add -u 2>/dev/null || true
+                git -C "$project_root" add -A 2>/dev/null || true
                 if git -C "$project_root" commit --no-verify -m "chore($feature_id): auto-commit session work" 2>/dev/null; then
                     log_info "Auto-commit succeeded"
                     session_status="success"
@@ -239,8 +239,8 @@ spawn_and_wait_session() {
             dirty_files=$(git -C "$project_root" status --porcelain 2>/dev/null || true)
             if [[ -n "$dirty_files" ]]; then
                 log_info "Auto-committing remaining session artifacts..."
-                git -C "$project_root" add -u 2>/dev/null || true
-                git -C "$project_root" commit --no-verify --amend --no-edit -a 2>/dev/null \
+                git -C "$project_root" add -A 2>/dev/null || true
+                git -C "$project_root" commit --no-verify --amend --no-edit 2>/dev/null \
                     || git -C "$project_root" commit --no-verify -m "chore($feature_id): include remaining session artifacts" 2>/dev/null \
                     || true
             fi
@@ -249,6 +249,9 @@ spawn_and_wait_session() {
 
     log_info "Session result: $session_status"
 
+    # Subagent detection
+    prizm_detect_subagents "$session_log"
+
     # Write lightweight session summary for post-session inspection
     local feature_slug
     feature_slug=$(python3 -c "

package/bundled/skills/_metadata.json

@@ -1,5 +1,5 @@
 {
-  "version": "1.0.126",
+  "version": "1.0.128",
   "skills": {
     "prizm-kit": {
       "description": "Full-lifecycle dev toolkit. Covers spec-driven development, Prizm context docs, code quality, debugging, deployment, and knowledge management.",
@@ -328,12 +328,15 @@
         ]
       },
       {
-        "name": "ui-ux-pro-max",
-        "description": "UI/UX design review and suggestions",
-        "repo": "https://github.com/nextlevelbuilder/ui-ux-pro-max-skill",
+        "name": "impeccable",
+        "description": "Production-grade UI/UX design skills (21 skills: frontend-design, adapt, animate, polish, etc.)",
+        "repo": "pbakaus/impeccable",
+        "installAll": true,
         "tags": [
           "design",
-          "frontend"
+          "frontend",
+          "ui",
+          "ux"
         ]
       }
     ]

package/bundled/skills/bugfix-pipeline-launcher/SKILL.md

@@ -111,61 +111,100 @@ Detect user intent from their message, then follow the corresponding workflow:
      --action status 2>/dev/null
    ```
 
-4. **Ask execution mode**: Present the user with a choice before launching:
-   - **(1) Foreground in session (recommended)**: Pipeline runs in the current session via `run-bugfix.sh run`. Visible output and direct error feedback.
-   - **(2) Background daemon**: Pipeline runs fully detached via `launch-bugfix-daemon.sh`. Survives session closure. Use only when user explicitly requests background execution.
-   - **(3) Manual — show commands**: Display the exact commands the user can run themselves. No execution.
+4. **Present all runtime options**: Show the user a complete options summary before launching. Present each option with its default, so the user can confirm or override.
 
-   Default to option 1 if user says "just run it" or doesn't specify. Default to option 2 only if user explicitly says "background", "detached", or "run in background".
+   **Runtime Options:**
 
-   **If option 1 (foreground)**:
+   | Option | Default | Description |
+   |--------|---------|-------------|
+   | **Execution mode** | Foreground | (1) Foreground — visible output (2) Background daemon — survives session closure (3) Manual — show commands only |
+   | **Verbose logging** | On | Detailed AI session logs including tool calls and subagent activity |
+   | **Max retries** | 3 | Max retry attempts per failed bug |
+   | **Session timeout** | None | Per-bug timeout in seconds (e.g. `3600` = 1 hour) |
+   | **Bug filter** | All | Run specific bugs: `B-001:B-005` (range), `B-001,B-003` (list), or mixed `B-001,B-005:B-010` |
+
+   Default to Foreground + Verbose On if user doesn't specify.
+
+   **Environment variable mapping** (for natural language → env var translation):
+
+   | User says | Environment variable |
+   |-----------|---------------------|
+   | "timeout 2 hours" | `SESSION_TIMEOUT=7200` |
+   | "max 5 retries" | `MAX_RETRIES=5` |
+   | "no verbose" / "quiet" | `VERBOSE=0` |
+   | "heartbeat every 60s" | `HEARTBEAT_INTERVAL=60` |
+
+   Example presentation to user:
+   ```
+   Bugfix pipeline will process N bugs with these settings:
+   - Execution:   Foreground (recommended)
+   - Verbose:     On (subagent detection enabled)
+   - Max retries: 3
+   - Timeout:     none
+   - Bugs:        all (by severity order)
+
+   Want to change any options, or launch with these defaults?
+   ```
+
+   **Execution mode details:**
+
+   **Foreground (recommended)**: Pipeline runs in the current session via `run-bugfix.sh run`. Provides visible output and direct error feedback.
+   ```bash
+   VERBOSE=1 dev-pipeline/run-bugfix.sh run bug-fix-list.json
+   ```
+   With all options:
    ```bash
-   dev-pipeline/run-bugfix.sh run bug-fix-list.json
+   VERBOSE=1 MAX_RETRIES=5 SESSION_TIMEOUT=3600 \
+     dev-pipeline/run-bugfix.sh run bug-fix-list.json
    ```
 
-   **If option 2 (background)**:
+   **Background daemon**: Pipeline runs fully detached via `launch-bugfix-daemon.sh`. Survives session closure. Use only when user explicitly requests background execution.
    ```bash
-   dev-pipeline/launch-bugfix-daemon.sh start bug-fix-list.json
+   dev-pipeline/launch-bugfix-daemon.sh start bug-fix-list.json --env "VERBOSE=1"
    ```
-   After launch, **record the decision** for traceability:
+   With all options:
    ```bash
-   echo "[$(date -u +%Y-%m-%dT%H:%M:%S)] MODE=daemon PID=$(cat dev-pipeline/bugfix-state/daemon.pid 2>/dev/null) BUG_LIST=bug-fix-list.json BUGS=$(python3 -c "import json; print(len(json.load(open('bug-fix-list.json')).get('bugs',[])))")" >> .prizmkit/bugfix-pipeline-run.log
+   dev-pipeline/launch-bugfix-daemon.sh start bug-fix-list.json \
+     --env "VERBOSE=1 MAX_RETRIES=5"
    ```
-   Note: Pipeline runs fully detached. Survives session closure.
 
-   **If option 3 (manual)**: Print commands and stop. Do not execute anything.
+   **Manual — show commands**: Display the exact commands the user can run themselves. Print commands and stop. Do not execute anything.
    ```
    # To run in foreground (recommended):
-   dev-pipeline/run-bugfix.sh run bug-fix-list.json
+   VERBOSE=1 dev-pipeline/run-bugfix.sh run bug-fix-list.json
 
    # To run in background (detached):
-   dev-pipeline/launch-bugfix-daemon.sh start bug-fix-list.json
+   dev-pipeline/launch-bugfix-daemon.sh start bug-fix-list.json --env "VERBOSE=1"
 
    # To check status:
    dev-pipeline/run-bugfix.sh status bug-fix-list.json
    ```
 
-5. **Ask user to confirm**: "Ready to launch the bugfix pipeline? It will process N bugs (by severity order)."
+5. **Ask user to confirm**: "Ready to launch the bugfix pipeline with the above settings?"
 
-6. **Launch** (based on chosen mode — see step 4).
+6. **Launch** with the assembled command from step 4.
 
-7. **Verify launch**:
-   ```bash
-   dev-pipeline/launch-bugfix-daemon.sh status
-   ```
+7. **Post-launch** (depends on execution mode):
 
-8. **Start log monitoring** (daemon mode only) -- Use the Bash tool with `run_in_background: true`:
-   ```bash
-   tail -f dev-pipeline/bugfix-state/pipeline-daemon.log
-   ```
-   This runs in background so you can continue interacting with the user.
+   **If foreground**: Pipeline runs to completion in the terminal. After it finishes:
+   - Summarize results: total bugs, fixed, failed, skipped
+   - If all fixed: each bug session has already run `prizmkit-retrospective` (structural sync) internally. Ask user what's next.
+   - If some failed: show failed bug IDs and suggest `retry-bug.sh <B-XXX>` or `reset-bug.sh <B-XXX> --clean --run`
 
-9. **Report to user**:
-   - Execution mode chosen
-   - Pipeline PID (daemon mode) or session status (foreground)
-   - Log file location
-   - "You can ask me 'bugfix status' or 'show fix logs' at any time"
-   - (daemon mode only) "Closing this session will NOT stop the pipeline"
+   **If background daemon**:
+   1. Verify launch:
+      ```bash
+      dev-pipeline/launch-bugfix-daemon.sh status
+      ```
+   2. Start log monitoring — Use the Bash tool with `run_in_background: true`:
+      ```bash
+      tail -f dev-pipeline/bugfix-state/pipeline-daemon.log
+      ```
+   3. Report to user:
+      - Pipeline PID
+      - Log file location
+      - "You can ask me 'bugfix status' or 'show fix logs' at any time"
+      - "Closing this session will NOT stop the pipeline"
 
 ---
 
@@ -236,25 +275,7 @@ Detect user intent from their message, then follow the corresponding workflow:
 
 ---
 
-#### Intent E: Custom Parameters
-
-When user specifies custom settings, map to environment variables:
-
-| User says | Environment variable |
-|-----------|---------------------|
-| "timeout 2 hours" | `SESSION_TIMEOUT=7200` |
-| "max 5 retries" | `MAX_RETRIES=5` |
-| "verbose mode" | `VERBOSE=1` |
-| "heartbeat every 60s" | `HEARTBEAT_INTERVAL=60` |
-
-Pass via `--env`:
-```bash
-dev-pipeline/launch-bugfix-daemon.sh start bug-fix-list.json --env "SESSION_TIMEOUT=7200 MAX_RETRIES=5 VERBOSE=1"
-```
-
----
-
-#### Intent F: Retry Single Bug
+#### Intent E: Retry Single Bug
 
 When user says "retry B-001":
 

package/bundled/skills/dev-pipeline-launcher/SKILL.md

@@ -105,80 +105,103 @@ Detect user intent from their message, then follow the corresponding workflow:
      --action status 2>/dev/null
    ```
 
-4. **Ask execution mode**: Present the user with a choice before launching:
-   - **(1) Foreground in session (recommended)**: Pipeline runs in the current session via `run.sh run`. Visible output and direct error feedback.
-   - **(2) Background daemon**: Pipeline runs fully detached via `launch-daemon.sh`. Survives session closure. Use only when user explicitly requests background execution.
-   - **(3) Manual — show commands**: Display the exact commands the user can run themselves. No execution.
+4. **Present all runtime options**: Show the user a complete options summary before launching. Present each option with its default, so the user can confirm or override.
 
-   Default to option 1 if user says "just run it" or doesn't specify. Default to option 2 only if user explicitly says "background", "detached", or "run in background".
+   **Runtime Options:**
 
-   **If option 1 (foreground)**:
+   | Option | Default | Description |
+   |--------|---------|-------------|
+   | **Execution mode** | Foreground | (1) Foreground — visible output (2) Background daemon — survives session closure (3) Manual — show commands only |
+   | **Critic review** | Off | Adversarial review after planning & implementation. Increases time ~5-10 min/feature |
+   | **Verbose logging** | On | Detailed AI session logs including tool calls and subagent activity |
+   | **Max retries** | 3 | Max retry attempts per failed feature |
+   | **Session timeout** | None | Per-feature timeout in seconds (e.g. `3600` = 1 hour) |
+   | **Feature filter** | All | Run specific features: `F-001:F-005` (range), `F-001,F-003` (list), or mixed `F-001,F-005:F-010` |
+
+   Default to Foreground + Verbose On if user doesn't specify. Default Critic to Off unless features have `estimated_complexity: "high"` or above.
+
+   **Environment variable mapping** (for natural language → env var translation):
+
+   | User says | Environment variable |
+   |-----------|---------------------|
+   | "timeout 2 hours" | `SESSION_TIMEOUT=7200` |
+   | "max 5 retries" | `MAX_RETRIES=5` |
+   | "no verbose" / "quiet" | `VERBOSE=0` |
+   | "heartbeat every 60s" | `HEARTBEAT_INTERVAL=60` |
+   | "enable critic review" | `ENABLE_CRITIC=true` |
+
+   Example presentation to user:
+   ```
+   Pipeline will process N features with these settings:
+   - Execution:   Foreground (recommended)
+   - Critic:      Off
+   - Verbose:     On (subagent detection enabled)
+   - Max retries: 3
+   - Timeout:     none
+   - Features:    all
+
+   Want to change any options, or launch with these defaults?
+   ```
+
+   **Execution mode details:**
+
+   **Foreground (recommended)**: Pipeline runs in the current session via `run.sh run`. Provides visible output and direct error feedback.
    ```bash
-   dev-pipeline/run.sh run feature-list.json
+   VERBOSE=1 dev-pipeline/run.sh run feature-list.json
    ```
-   If user wants to run only specific features:
+   With all options:
    ```bash
-   dev-pipeline/run.sh run feature-list.json --features F-001:F-005
-   dev-pipeline/run.sh run feature-list.json --features F-001,F-003,F-007
+   VERBOSE=1 ENABLE_CRITIC=true MAX_RETRIES=5 SESSION_TIMEOUT=3600 \
+     dev-pipeline/run.sh run feature-list.json --features F-001:F-005
    ```
 
-   **If option 2 (background)**:
+   **Background daemon**: Pipeline runs fully detached via `launch-daemon.sh`. Survives session closure. Use only when user explicitly requests background execution.
    ```bash
-   dev-pipeline/launch-daemon.sh start feature-list.json
+   dev-pipeline/launch-daemon.sh start feature-list.json --env "VERBOSE=1"
    ```
-   With feature filter:
+   With all options:
    ```bash
-   dev-pipeline/launch-daemon.sh start feature-list.json --features F-001:F-005
+   dev-pipeline/launch-daemon.sh start feature-list.json --features F-001:F-005 \
+     --env "VERBOSE=1 ENABLE_CRITIC=true MAX_RETRIES=5"
    ```
-   Note: Pipeline runs fully detached. Survives session closure.
 
-   **If option 3 (manual)**: Print commands and stop. Do not execute anything.
+   **Manual — show commands**: Display the exact commands the user can run themselves. Print commands and stop. Do not execute anything.
    ```
    # To run in foreground (recommended):
-   dev-pipeline/run.sh run feature-list.json
+   VERBOSE=1 dev-pipeline/run.sh run feature-list.json
 
    # To run in background (detached):
-   dev-pipeline/launch-daemon.sh start feature-list.json
+   dev-pipeline/launch-daemon.sh start feature-list.json --env "VERBOSE=1"
 
    # To check status:
    dev-pipeline/run.sh status feature-list.json
    ```
 
-5. **Ask whether to enable Critic Agent** (adversarial review):
-   Present the choice:
-   - **(a) No — standard pipeline (default)**: No adversarial review. Faster execution, lower token cost.
-   - **(b) Yes — enable Critic review**: Adds adversarial challenge after planning and implementation. Challenges plan fitness and code integration quality. Increases pipeline time by ~5-10 minutes per feature.
-
-   Default to (a). Only suggest (b) if features have `estimated_complexity: "high"` or above.
-
-   If user chooses (b), prepend `ENABLE_CRITIC=true` to the launch command in step 6.
-
-6. **Ask user to confirm**: "Ready to launch the pipeline? It will process N features."
+5. **Ask user to confirm**: "Ready to launch the pipeline with the above settings?"
 
-7. **Launch** (based on chosen mode from step 4, with `ENABLE_CRITIC=true` env var if chosen in step 5):
-   - Foreground: `ENABLE_CRITIC=true dev-pipeline/run.sh run feature-list.json`
-   - Background: `dev-pipeline/launch-daemon.sh start feature-list.json --env "ENABLE_CRITIC=true"`
-   - If user specified environment overrides:
-     ```bash
-     dev-pipeline/launch-daemon.sh start feature-list.json --env "SESSION_TIMEOUT=7200 MAX_RETRIES=5"
-     ```
+6. **Launch** with the assembled command from step 4.
 
-8. **Verify launch**:
-   ```bash
-   dev-pipeline/launch-daemon.sh status
-   ```
+7. **Post-launch** (depends on execution mode):
 
-9. **Start log monitoring** -- Use the Bash tool with `run_in_background: true`:
-   ```bash
-   tail -f dev-pipeline/state/pipeline-daemon.log
-   ```
-   This runs in background so you can continue interacting with the user.
+   **If foreground**: Pipeline runs to completion in the terminal. After it finishes:
+   - Summarize results: total features, succeeded, failed, skipped
+   - If all succeeded: each feature session has already run `prizmkit-retrospective` internally. Ask user what's next.
+   - If some failed: show failed feature IDs and suggest `retry-feature.sh <F-XXX>` or `reset-feature.sh <F-XXX> --clean --run`
 
-10. **Report to user**:
-   - Pipeline PID
-   - Log file location
-   - "You can ask me 'pipeline status' or 'show logs' at any time"
-   - "Closing this session will NOT stop the pipeline"
+   **If background daemon**:
+   1. Verify launch:
+      ```bash
+      dev-pipeline/launch-daemon.sh status
+      ```
+   2. Start log monitoring — Use the Bash tool with `run_in_background: true`:
+      ```bash
+      tail -f dev-pipeline/state/pipeline-daemon.log
+      ```
+   3. Report to user:
+      - Pipeline PID
+      - Log file location
+      - "You can ask me 'pipeline status' or 'show logs' at any time"
+      - "Closing this session will NOT stop the pipeline"
 
 ---
 
@@ -249,54 +272,7 @@ Detect user intent from their message, then follow the corresponding workflow:
 
 ---
 
-#### Intent E: Custom Parameters
-
-When user specifies custom settings, map to environment variables:
-
-| User says | Environment variable |
-|-----------|---------------------|
-| "timeout 2 hours" | `SESSION_TIMEOUT=7200` |
-| "max 5 retries" | `MAX_RETRIES=5` |
-| "verbose mode" | `VERBOSE=1` |
-| "heartbeat every 60s" | `HEARTBEAT_INTERVAL=60` |
-| "enable critic review" | `ENABLE_CRITIC=true` env var (or `--critic` for single feature) |
-
-Pass via `--env`:
-```bash
-dev-pipeline/launch-daemon.sh start feature-list.json --env "SESSION_TIMEOUT=7200 MAX_RETRIES=5 VERBOSE=1"
-```
-
----
-
-#### Intent F: Run Subset of Features
-
-When user says "run only F-001 to F-005", "only build features 1 through 5":
-
-```bash
-dev-pipeline/run.sh run feature-list.json --features F-001:F-005
-```
-
-When user says "run F-001, F-003, and F-007", "only build these three features":
-
-```bash
-dev-pipeline/run.sh run feature-list.json --features F-001,F-003,F-007
-```
-
-Mixed format (IDs + ranges):
-
-```bash
-dev-pipeline/run.sh run feature-list.json --features F-001,F-005:F-010
-```
-
-Background daemon with feature filter:
-
-```bash
-dev-pipeline/launch-daemon.sh start feature-list.json --features F-001:F-005
-```
-
----
-
-#### Intent G: Retry Single Feature Node
+#### Intent E: Retry Single Feature Node
 
 When user says "retry F-003":
 
@@ -341,4 +317,4 @@ Notes:
 - **Single instance**: Only one pipeline can run at a time. The PID file prevents duplicates.
 - **Pipeline coexistence**: Feature and bugfix pipelines use separate state directories (`state/` vs `bugfix-state/`), so they can run simultaneously without conflict.
 - **State preservation**: Stopping and restarting the pipeline resumes from where it left off -- completed features are not re-run.
-- **HANDOFF**: After pipeline completes all features, suggest running `prizmkit-retrospective` for `.prizm-docs/` architecture sync, or ask user what's next.
+- **HANDOFF**: After pipeline completes all features, each session has already run `prizmkit-retrospective` internally. Ask user what's next.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "prizmkit",
-  "version": "1.0.126",
+  "version": "1.0.128",
   "description": "Create a new PrizmKit-powered project with clean initialization — no framework dev files, just what you need.",
   "type": "module",
   "bin": {

package/src/external-skills.js

@@ -1,11 +1,14 @@
 /**
  * External Skills Installer
  *
- * Delegates to `npx skills add <repo> --skill <name> --yes` to fetch the skill.
- * When run non-interactively (stdio: 'pipe'), npx skills only writes the canonical
- * .agents/skills/<name>/SKILL.md without creating any platform symlinks.
- * We read that file, copy it to the selected platform dir(s), then clean up
- * .agents/ and skills-lock.json.
+ * Delegates to `npx skills add <repo> [--skill <name>] --yes` to fetch skills.
+ * Supports two modes:
+ *   - Single skill: `--skill <name>` selects one skill from a repo
+ *   - Install all:  installs all skills from a repo (e.g. pbakaus/impeccable)
+ *
+ * When run non-interactively (stdio: 'pipe'), npx skills writes canonical
+ * .agents/skills/<name>/SKILL.md files. We copy those to the selected
+ * platform dir(s), then clean up .agents/ and skills-lock.json.
  */
 
 import { execSync } from 'node:child_process';
@@ -14,8 +17,9 @@ import fs from 'fs-extra';
 
 /**
  * @param {Object} skill        - Skill definition from external_skills.known
- *   skill.repo  {string}       - GitHub URL
- *   skill.name  {string}       - Skill name as listed in the repo
+ *   skill.repo       {string}  - GitHub short ref (e.g. "pbakaus/impeccable") or full URL
+ *   skill.name       {string}  - Skill name (used for display / single-skill mode)
+ *   skill.installAll {boolean} - If true, install all skills from the repo
  * @param {string} platform     - 'claude' | 'codebuddy' | 'both'
  * @param {string} projectRoot  - Target project root
  * @param {boolean} dryRun
@@ -25,35 +29,60 @@ export async function installExternalSkill(skill, platform, projectRoot, dryRun)
     throw new Error(`Skill "${skill.name}" has no repo URL defined`);
   }
 
-  const cmd = `npx skills add ${skill.repo} --skill ${skill.name} --yes`;
+  // installAll: install entire repo (all skills), otherwise install single skill by name
+  const cmd = skill.installAll
+    ? `npx skills add ${skill.repo} --yes`
+    : `npx skills add ${skill.repo} --skill ${skill.name} --yes`;
 
   if (dryRun) {
-    const targets = getTargetDirs(platform, projectRoot)
-      .map(d => path.relative(projectRoot, path.join(d, skill.name, 'SKILL.md')));
     console.log(`      [dry-run] ${cmd}`);
-    for (const t of targets) console.log(`      [dry-run] → ${t}`);
+    if (skill.installAll) {
+      console.log(`      [dry-run] → all skills from ${skill.repo}`);
+    } else {
+      const targets = getTargetDirs(platform, projectRoot)
+        .map(d => path.relative(projectRoot, path.join(d, skill.name, 'SKILL.md')));
+      for (const t of targets) console.log(`      [dry-run] → ${t}`);
+    }
     return;
   }
 
-  // stdio: 'pipe' → npx skills runs non-interactively, writes only .agents/skills/<name>/SKILL.md
-  execSync(cmd, { cwd: projectRoot, stdio: 'pipe', timeout: 60000 });
+  // stdio: 'pipe' → npx skills runs non-interactively
+  execSync(cmd, { cwd: projectRoot, stdio: 'pipe', timeout: 120000 });
 
-  const canonicalSkillMd = path.join(projectRoot, '.agents', 'skills', skill.name, 'SKILL.md');
-  if (!await fs.pathExists(canonicalSkillMd)) {
-    throw new Error(`npx skills did not produce .agents/skills/${skill.name}/SKILL.md`);
-  }
+  const agentsSkillsDir = path.join(projectRoot, '.agents', 'skills');
+
+  if (skill.installAll) {
+    // Repo-level install: copy ALL skill dirs from .agents/skills/ to platform dirs
+    if (!await fs.pathExists(agentsSkillsDir)) {
+      throw new Error(`npx skills did not produce .agents/skills/ for repo ${skill.repo}`);
+    }
+    const skillDirs = (await fs.readdir(agentsSkillsDir, { withFileTypes: true }))
+      .filter(d => d.isDirectory())
+      .map(d => d.name);
+
+    for (const targetDir of getTargetDirs(platform, projectRoot)) {
+      for (const skillName of skillDirs) {
+        const srcDir = path.join(agentsSkillsDir, skillName);
+        const destDir = path.join(targetDir, skillName);
+        await fs.remove(destDir).catch(() => {});
+        await fs.copy(srcDir, destDir);
+      }
+    }
+  } else {
+    // Single-skill install
+    const canonicalSkillMd = path.join(agentsSkillsDir, skill.name, 'SKILL.md');
+    if (!await fs.pathExists(canonicalSkillMd)) {
+      throw new Error(`npx skills did not produce .agents/skills/${skill.name}/SKILL.md`);
+    }
 
-  const content = await fs.readFile(canonicalSkillMd, 'utf8');
-
-  // Copy to selected platform dir(s)
-  // npx skills may have created symlinks (e.g. .claude/skills/<name> -> ../../.agents/skills/<name>)
-  // We must remove those symlinks first so we can write a real directory with the actual content.
-  for (const targetDir of getTargetDirs(platform, projectRoot)) {
-    const skillDir = path.join(targetDir, skill.name);
-    // Remove any existing symlink or directory at this path before creating a real dir
-    await fs.remove(skillDir).catch(() => {});
-    await fs.ensureDir(skillDir);
-    await fs.writeFile(path.join(skillDir, 'SKILL.md'), content);
+    const content = await fs.readFile(canonicalSkillMd, 'utf8');
+
+    for (const targetDir of getTargetDirs(platform, projectRoot)) {
+      const skillDir = path.join(targetDir, skill.name);
+      await fs.remove(skillDir).catch(() => {});
+      await fs.ensureDir(skillDir);
+      await fs.writeFile(path.join(skillDir, 'SKILL.md'), content);
+    }
   }
 
   // Clean up npx skills artifacts after copying