Merge dev → main (automated)

.githooks/pre-push

@@ -19,6 +19,22 @@ if [ "$WORKLOG_SKIP_PRE_PUSH" = "1" ]; then
   exit 0
 fi
 
+# Skip when running inside a temp worktree created by withTempWorktree
+# for internal sync operations. These worktrees don't have worklog
+# initialized and the push is already done with --no-verify as defense.
+case "$PWD" in
+  *tmp-worktree-*)
+    exit 0
+    ;;
+esac
+
+# Skip when inside a git worktree (not the main checkout).
+# Worktrees are for feature development and typically don't have worklog
+# initialized. The sync should run from the main checkout.
+if [ "$(git rev-parse --git-dir 2>/dev/null)" != "$(git rev-parse --git-common-dir 2>/dev/null)" ]; then
+  exit 0
+fi
+
 # Read stdin to check which refs are being pushed.
 # If pushing to refs/worklog/data, skip the sync to avoid infinite loops.
 skip=0
@@ -44,5 +60,8 @@ fi
 # Force the data branch to refs/worklog/data regardless of config.
 # This prevents the sync from accidentally pushing to a standard branch
 # if the user has overridden syncBranch in their worklog config.
-"$WL" sync --git-branch refs/worklog/data
+"$WL" sync --git-branch refs/worklog/data || {
+  echo "worklog: pre-push sync failed (pushing anyway)" >&2
+  exit 0
+}
 exit 0

.github/workflows/install-and-smoke-test.yml

@@ -19,7 +19,7 @@ jobs:
       - name: Setup Node
         uses: actions/setup-node@v4
         with:
-          node-version: "20"
+          node-version: "22"
           cache: "npm"
 
       - name: Install dependencies
@@ -31,6 +31,11 @@ jobs:
       - name: Verify package builds
         run: npm pack
 
+      - name: Initialize Worklog
+        run: |
+          node ./dist/cli.js init --project-name "ContextHub" --prefix WL --agents-template skip --json > /dev/null
+          echo "✓ Worklog initialized"
+
       - name: Run headless smoke test
         run: |
           # Verify the wl command is available and works
@@ -39,30 +44,36 @@ jobs:
 
           # Verify the TUI module loads without errors (using tsx for TypeScript imports)
           npx tsx -e "
-            const { ChatPane } = await import('./packages/tui/extensions/chatPane.js');
-            const { ActionPalette } = await import('./packages/tui/extensions/actionPalette.js');
-            const { runWl } = await import('./packages/tui/extensions/wl-integration.js');
+          (async () => {
+            const { ChatPane } = await import('./packages/tui/extensions/Worklog/chatPane.ts');
+            const { ActionPalette } = await import('./packages/tui/extensions/Worklog/actionPalette.ts');
+            const { runWl } = await import('./packages/tui/extensions/wl-integration.ts');
             console.log('✓ TUI modules load successfully');
+          })();
           "
 
           # Verify chat pane can send messages
           npx tsx -e "
-            const { ChatPane } = await import('./packages/tui/extensions/chatPane.js');
+          (async () => {
+            const { ChatPane } = await import('./packages/tui/extensions/Worklog/chatPane.ts');
             const pane = new ChatPane();
             pane.clear();
             console.log('✓ ChatPane instantiated');
+          })();
           "
 
           # Verify action palette has default actions
           npx tsx -e "
-            const { ChatPane } = await import('./packages/tui/extensions/chatPane.js');
-            const { ActionPalette } = await import('./packages/tui/extensions/actionPalette.js');
+          (async () => {
+            const { ChatPane } = await import('./packages/tui/extensions/Worklog/chatPane.ts');
+            const { ActionPalette } = await import('./packages/tui/extensions/Worklog/actionPalette.ts');
             const chat = new ChatPane();
             const palette = new ActionPalette(chat);
             palette.open();
             const actions = palette.getFilteredActions();
             if (actions.length < 5) throw new Error('Expected at least 5 default actions, got ' + actions.length);
             console.log('✓ ActionPalette has ' + actions.length + ' default actions');
+          })();
           "
 
           # Verify pi-audit module works

.github/workflows/tui-tests.yml

@@ -13,7 +13,6 @@ jobs:
     outputs:
       cli: ${{ steps.filter.outputs.cli }}
       shared: ${{ steps.filter.outputs.shared }}
-      tui: ${{ steps.filter.outputs.tui }}
     steps:
       - name: Checkout
         uses: actions/checkout@v4
@@ -38,27 +37,16 @@ jobs:
               - 'src/github*.ts'
               - 'src/types.ts'
               - 'src/index.ts'
-              - 'tests/cli.test.ts'
               - 'tests/**/*.test.ts'
               - 'test/**/*.test.ts'
-              - 'tests/e2e/**'
-            tui:
-              - 'src/commands/tui.ts'
-              - 'src/tui/**'
-              - 'tests/tui/**/*.test.ts'
-              - 'test/tui-*.test.ts'
-              - 'tests/tui-ci-run.sh'
-              - 'vitest.tui.config.ts'
-              - 'Dockerfile.tui-tests'
-              - 'test-tui.sh'
 
   docs-only:
     runs-on: ubuntu-latest
     needs: changes
-    if: ${{ needs.changes.outputs.cli != 'true' && needs.changes.outputs.tui != 'true' && needs.changes.outputs.shared != 'true' }}
+    if: ${{ needs.changes.outputs.cli != 'true' && needs.changes.outputs.shared != 'true' }}
     steps:
       - name: No test changes detected
-        run: echo "Docs-only change; skipping CLI/TUI test jobs."
+        run: echo "Docs-only change; skipping CLI test jobs."
 
   cli-tests:
     runs-on: ubuntu-latest
@@ -71,7 +59,7 @@ jobs:
       - name: Setup Node
         uses: actions/setup-node@v4
         with:
-          node-version: "20"
+          node-version: "22"
           cache: "npm"
 
       - name: Install dependencies
@@ -80,25 +68,10 @@ jobs:
       - name: Build CLI
         run: npm run build
 
+      - name: Initialize Worklog
+        run: |
+          node ./dist/cli.js init --project-name "ContextHub" --prefix WL --agents-template skip --json > /dev/null
+          echo "✓ Worklog initialized"
+
       - name: Run CLI tests
         run: npm test
-
-  tui-tests:
-    runs-on: ubuntu-latest
-    needs: changes
-    if: ${{ needs.changes.outputs.tui == 'true' || needs.changes.outputs.shared == 'true' }}
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Setup Node
-        uses: actions/setup-node@v4
-        with:
-          node-version: "20"
-          cache: "npm"
-
-      - name: Install dependencies
-        run: npm ci
-
-      - name: Run headless TUI tests
-        run: bash ./tests/tui-ci-run.sh

.implement_state.json

@@ -0,0 +1,8 @@
+{
+  "work_item_id": "WL-0MRSGMTYU009OZCF",
+  "worktree_path": "/home/rgardler/projects/ContextHub/.worklog/worktrees/wl-WL-0MRSGMTYU009OZCF-add-audit-file-to-wl-audit-set-for-file",
+  "repo_root": "/home/rgardler/projects/ContextHub",
+  "parent_branch": "dev",
+  "commit_msg": "",
+  "started_at": "2026-07-21T18:30:22Z"
+}

.pi/settings.json

@@ -5,6 +5,7 @@
   "context-hub": {
     "showActivityIndicator": true,
     "showHelpText": true,
-    "browseItemCount": 15
+    "browseItemCount": 15,
+    "schedules": []
   }
 }

.pi/skills/heartbeat/SKILL.md

@@ -0,0 +1,144 @@
+---
+name: heartbeat
+description: "Automated work item monitoring — inspects the completed/in_review queue and either flags the next item for producer review (when sparse) or runs an audit on the first unverified item (when full). Trigger on queries like: '/skill:heartbeat'"
+---
+
+# Heartbeat Skill
+
+## Completion-Detection Gate
+
+The purpose of this gate is to prevent the heartbeat from interrupting work
+in progress. However, the gate logic differs depending on **who** is invoking
+the skill.
+
+### User invocation (`/skill:heartbeat`)
+
+When the skill is invoked by a **user** via `/skill:heartbeat`, **proceed
+immediately**. The user is taking explicit control and does not need the gate.
+
+This is equivalent to passing `--force` to the script — no conversation-history
+check is performed.
+
+### Agent-initiated invocation
+
+When the agent invokes the heartbeat **autonomously** (e.g., in an idle loop,
+cron-like context, or without an explicit user request), check whether the
+previous work has completed to avoid interrupting mid-process tasks:
+
+1. Review the conversation history up to (but not including) this
+   invocation. Find the **last assistant message**
+   (the most recent message from the Pi agent before this turn).
+2. If that message **clearly states that a process has completed**
+   (e.g., "Work committed to dev", "Task complete", "All tests pass",
+   a work-item summary with "Work committed to dev", or any message that
+   unambiguously signals the end of a task), **proceed** to the heartbeat
+   logic below.
+3. If the last assistant message **does not clearly indicate completion**
+   (e.g., it asks a question, reports an error, or is mid-process),
+   print the following and exit **without** taking any heartbeat action:
+
+   ```
+   No completed process detected — heartbeat taking no action.
+   ```
+
+   Do NOT run `./scripts/heartbeat.py` in this case.
+
+4. **Direct script invocation bypass:** When the script is run directly
+   (outside Pi, e.g., from CI or a terminal), there is no agent to perform
+   the gate check. In that case, pass `--force` to the script to bypass:
+
+   ```bash
+   python3 skill/heartbeat/scripts/heartbeat.py --force
+   ```
+
+   The `--force` flag has no effect inside the script (the gate is an
+   agent-level concern); it is a documentation/handshake flag indicating
+   that the caller takes responsibility for the gate check.
+
+## Overview
+
+The heartbeat skill automates work item queue monitoring. It inspects the
+completed/in_review queue and takes one of three actions depending on queue
+state:
+
+- **Sparse queue (< 10 items):** Finds the next ready work item via `wl next`
+  and flags it for producer review (`needsProducerReview: true`).
+- **Full queue (>= 10 items):** Finds the first item (by `sortIndex`) that
+  does **not** have a valid audit result ("Ready to close: Yes") and runs
+  `/skill:audit <id>` via the Pi framework on it.
+- **All items ready:** Reports "Project is ready for producer review prior to
+  a new release".
+
+Only one audit is triggered per heartbeat invocation.
+
+## Invocation
+
+```
+/skill:heartbeat
+```
+
+Invoke via the Pi chat interface. Output is displayed directly in the chat.
+
+## Behavior
+
+After passing the completion-detection gate above:
+
+1. Run `python3 ./scripts/heartbeat.py` (or `python3 [--force]` for
+   standalone/automated use).
+2. The script queries `wl list --status completed --stage in_review --json`
+   to count items.
+3. **If count < 10 (sparse queue):**
+   - Call `wl next --json` to find the next ready work item.
+   - If found, set `needsProducerReview: true` via `wl update <id> --needs-producer-review true`.
+   - Report which item was flagged.
+   - If no item returned by `wl next`, report gracefully.
+4. **If count >= 10 (full queue):**
+   - Sort items by `sortIndex` ascending.
+   - For each item, check for a valid audit result via `wl audit-show <id> --json`.
+   - An item is considered "ready" if its `rawOutput` starts with "Ready to close: Yes".
+   - Find the first non-ready item and run `/skill:audit <id>` via the Pi framework (`pi -p --mode json --no-session /skill:audit <id>`).
+   - Only one item is audited per invocation.
+5. **If all items have "Ready to close: Yes":**
+   - Report "Project is ready for producer review prior to a new release".
+
+## Inputs
+
+None (when invoked via `/skill:heartbeat`). The skill operates on the current
+state of the Worklog database.
+
+The underlying script accepts:
+
+- `--force` — Bypass the completion-detection gate. Use when running the
+  script standalone (e.g., from CI, cron, or a scheduler that already
+  performs its own idle/completion checks).
+
+## Outputs
+
+Human-readable text printed to stdout (displayed in the Pi chat interface).
+
+## Exit Codes
+
+- 0 — Success (action taken or all items ready)
+- 1 — Error (wl command failure, JSON parse error)
+
+## Dependencies
+
+- `wl` CLI (Worklog) — must be installed and in PATH
+- `pi` CLI — must be installed and in PATH (for invoking `/skill:audit`)
+- Audit skill — located at `~/.pi/agent/skills/audit/`
+- Python 3 — for running the heartbeat script
+
+## Key Files
+
+- `skill/heartbeat/SKILL.md` — This file
+- `skill/heartbeat/scripts/heartbeat.py` — Core implementation script
+- `tests/skill/heartbeat/test_heartbeat.py` — Unit tests
+
+## Related Work Items
+
+- WL-0MLQ0ZHQE0JBX8Y6 — TUI filter shortcut for needsProducerReview
+- WL-0MLGTWT4S1X4HDD9 — `--needs-producer-review` filter for `wl list`
+- WL-0MR6XG7RX008AF2W — Closing question in audit skill output
+- WL-0MLYTKTI20V31KYW — Structured audit report format
+- WL-0MM347F9D1EGKLSQ — sortIndex selection with batch mode
+- WL-0MRJATQJ900832IT — Completion-detection gate