Merge dev → main (automated)

.worklog/config.yaml

@@ -4,4 +4,3 @@ autoSync: false
 githubRepo: TheWizardsCode/ContextHub
 githubLabelPrefix: 'wl:'
 githubImportCreateNew: true
-openBrainEnabled: true

CHANGELOG.md

@@ -1,5 +1,37 @@
 # Changelog
 
+## v1.0.18 (2026-10-04)
+### Features
+- Recent dispatches are now grouped by time of day, so you can see when work happened. (WL-0MUMM9NED009TLL3)
+- When no questions are found, you now get a clear explanation of what's needed to clear the review flag. (WL-0MUH765F1003FKRK)
+- Review and close idle agent panes in your workspace, with your approval before anything closes. (WL-0MUJMXVPO0016DZM)
+- Failed tasks now wait before being retried, preventing wasted slots on repeated failures. (WL-0MUKYERLZ006ELL5)
+- Downtime work now always picks the top item from your Herdr list, so automated tasks match what you see. (WL-0MTK1ILM2009QYB2)
+- Automated work panes now open in your project's workspace, grouped by task. (WL-0MU321YK70035AYT)
+- Press r then p for Producer Review, or r then i to start an interview. (WL-0MU95SGEB006HQAM)
+- Automatically detects and safely fixes performance issues slowing down your machine. (WL-0MUJLCVRS0093EJJ)
+- Covered items now show a dimmed audit icon so you can tell they're already reviewed. (WL-0MUBVH8QG0020H9L)
+- Completed work can no longer ship without an audit unnoticed, with warnings and a report to catch gaps. (WL-0MUBVH9FV0027COG)
+- Press f d to review the 20 most recent auto-dispatched work items and their outcomes. (WL-0MUGLL9SS002E1D2)
+- Audits now run one at a time per machine, so they finish reliably instead of failing when several run at once. (WL-0MUIVE0YG000UVIA)
+- Completed agent sessions now close automatically, freeing up your machine's resources. (WL-0MUJL1NAH0042GOS)
+- Audit freshness now works consistently everywhere, so items won't appear fresh in one view and stale in another. (WL-0MUBVH7ZR009PP80)
+- Automated downtime work now appears in its own project tab, keeping projects separate. (WL-0MTRQT482001SNXC)
+- Adds automated checks to keep work item recovery reliable, so stalled tasks resume correctly. (WL-0MTF5WO4Q005YBVH)
+- The review queue banner now tells you whether to keep auditing or ship. (WL-0MUJQFXYC009CFRJ)
+### Bug Fixes
+- Interview questions now display in full, so you can see every option and note before answering. (WL-0MUKCGV3X0030W6K)
+- Stage filters now show in-progress work instead of hiding it. (WL-0MUIB7D30009KG00)
+- Automated tasks now open in the correct project workspace instead of a shared one. (WL-0MUR5FUWD00024XN)
+- The top info line now stays visible even when the help text wraps. (WL-0MTV979LK005YB1B)
+- Rejecting work items with an apostrophe in the reason now works correctly. (WL-0MU7KEX65004X0U9)
+- Long audit sessions no longer crash from a stale notification after a session reload. (WL-0MUIV50EJ007VH9B)
+- Shortcut commands with special characters now work correctly. (WL-0MUKH3DVA00488U0)
+- Fixed a test that was checking for an outdated keyboard shortcut hint. (WL-0MUKDNC7L000I1JO)
+### Other
+- Old debug logs are now cleaned up automatically, keeping the game running fast and smooth. (WL-0MUJL0PTS009MC8F)
+- Completing a work item no longer shows stray error messages in your terminal. (WL-0MTH28WDS008WMXA)
+
 ## v1.0.17 (2026-09-30)
 ### Features
 - Press r then i to start an interview on the selected work item. (WL-0MU95SGEB006HQAM)

CLI.md

@@ -1165,15 +1165,12 @@ wl plugins
 
 Worklog comes bundled with an example stats plugin installed.
 
-- `openbrain` — Manage OpenBrain submission queue (`status`, `resubmit`).
 - `stats` — Show custom work item statistics (example plugin provided in this repo).
  - `ampa` — AMPA plugin: manage AMPA containers and workspace tasks (start, stop, status, run, list, start-work, finish-work).
 
 Examples:
 
 ```sh
-wl openbrain status
-wl openbrain resubmit WL-ABC123
 wl ampa start                 # start AMPA services for this repo
 wl ampa status                # show AMPA service status
 wl ampa list                  # list available AMPA containers/tasks
@@ -1256,19 +1253,67 @@ Answers are persisted after each response, so an interrupted session can be
 resumed by re-running the command: only the questions that remain outstanding
 are asked. `needsProducerReview` is cleared once every question has an answer.
 
+#### When there are no questions to answer
+
+If no clarifying-questions section is found (or it contains no parseable
+questions) but the item is still flagged for producer review, the command
+explains what the producer must do to clear the flag. The explanation is
+produced by the local LLM proxy's `compact` model, prompted with the
+instruction *"Explain what the producer needs to do in order to remove the
+needsProducerReview flag"* plus item context (the description and the two most
+recent comments). When the LLM is disabled or unreachable the command falls
+back silently to the two most recent comments (`<author>: <first line>`).
+
+The explanation is advisory only — it never mutates the description.
+
+After printing the explanation the command asks
+`Clear the needsProducerReview flag? (y/N)` (default **No**); answering `y`
+clears the flag via the store, anything else leaves it flagged. This prompt
+only appears in the `noSection`/`noQuestions` cases; the **`allAnswered`
+path is unchanged** (`WL-0MU55UDBJ008DJ67`) and still clears the flag
+automatically without prompting. No prompt is shown — and the flag is never
+mutated — in `--json` mode or other non-interactive contexts.
+
 Options:
 
 - `--prefix <prefix>` — Operate on a specific prefix (optional).
+- `--json` — Non-interactive mode: prints JSON with
+  `needsProducerReview` (boolean), `producerReviewExplanation` (string or
+  `null`), `noSection`, `noQuestions`, `total`, `outstanding` and
+  `allAnswered`. Performs no prompts, no mutation and (with `--no-llm`) no
+  LLM call.
+- `--no-llm` — Disable the LLM explanation and use the comment fallback
+  (default: LLM on).
+- `--model <model>` — Override the chat model used for the explanation.
 
-> The command requires interactive (TTY) input; it cannot be used with
-> `--json`.
+> Without `--json`, the command requires interactive (TTY) input.
 
 Examples:
 
 ```sh
 wl interview WL-ABC123
+wl interview WL-ABC123 --no-llm
+wl --json interview WL-ABC123
+```
+
+#### `llm` configuration
+
+The chat provider used for the explanation is configured under the `llm`
+section of `.worklog/config.yaml`, mirroring `embedding`:
+
+```yaml
+llm:
+  baseUrl: http://192.168.0.199:8000/v1
+  model: compact
+  apiKey: optional-key
+  timeoutMs: 15000
 ```
 
+Defaults: `baseUrl` `http://192.168.0.199:8000/v1`, `model` `compact`,
+`timeoutMs` `15000`. Environment variables `LLM_BASE_URL`, `LLM_MODEL`,
+`LLM_API_KEY` and `LLM_TIMEOUT_MS` are used as fallbacks; config values take
+precedence over environment variables.
+
 ### `help` [command]
 
 Show help for a specific command.

LOCAL_LLM.md

@@ -210,6 +210,29 @@ be used by agents via standard provider configuration (e.g. Ollama
 $env:OPENCODE_SERVER_PORT = 51625
 ```
 
+### Direct LLM chat features (`llm.*` config)
+
+Some `wl` features call an OpenAI-compatible chat-completions endpoint
+directly — currently the `wl interview` producer-review explanation, which
+uses the `ChatClient` in `src/lib/llm.ts`. These are configured under the
+`llm` section of `.worklog/config.yaml`, parallel to `embedding`:
+
+```yaml
+llm:
+  baseUrl: http://192.168.0.199:8000/v1
+  model: compact
+  apiKey: optional-key   # omit for local providers that need no auth
+  timeoutMs: 15000
+```
+
+Defaults: `baseUrl` `http://192.168.0.199:8000/v1`, `model` `compact`,
+`timeoutMs` `15000`. Environment variables `LLM_BASE_URL`, `LLM_MODEL`,
+`LLM_API_KEY` and `LLM_TIMEOUT_MS` are used as fallbacks; config values take
+precedence over environment variables (matching the `embedding.*` precedence).
+Every call is bounded by the timeout and degrades silently to a non-LLM
+fallback on error. See [CLI.md](CLI.md#interview-options-id) for the
+`wl interview --no-llm` opt-out.
+
 ---
 
 ## Task routing guidance (what to run locally vs hosted)

README.md

@@ -116,7 +116,7 @@ You can get a lot of value from using Worklog as a memory for your agents. But y
 |----------|-------------|
 | [Heartbeat Skill](skill/heartbeat/SKILL.md) | Automated work item monitoring and audit orchestration for the Pi agent |
 | [PLUGIN_GUIDE.md](PLUGIN_GUIDE.md) | Plugin development guide and API reference |
-| [LOCAL_LLM.md](LOCAL_LLM.md) | Configure local LLM providers (Ollama, Foundry) |
+| [LOCAL_LLM.md](LOCAL_LLM.md) | Configure local LLM providers (Ollama, Foundry) and the direct `llm.*` chat config used by `wl interview` |
 | [MULTI_PROJECT_GUIDE.md](MULTI_PROJECT_GUIDE.md) | Multi-project setup with custom prefixes |
 | [API.md](API.md) | REST API endpoints and usage |
 | [docs/FILE_PATH_CONVENTION.md](docs/FILE_PATH_CONVENTION.md) | File path convention for work item descriptions |
@@ -134,7 +134,6 @@ You can get a lot of value from using Worklog as a memory for your agents. But y
 | [AGENTS.md](AGENTS.md) | AI agent onboarding and workflow instructions |
 | [tests/README.md](tests/README.md) | Test suite documentation |
 | [examples/README.md](examples/README.md) | Example plugins |
- | [OpenBrain Integration](docs/openbrain.md) | Documentation for the optional OpenBrain submission integration |
 
 ### Internal / Development
 
@@ -148,6 +147,7 @@ You can get a lot of value from using Worklog as a memory for your agents. But y
 | [docs/dev/wl-process-healthcheck.md](docs/dev/wl-process-healthcheck.md) | wl process healthcheck watchdog usage (cron/systemd, thresholds) |
 | [docs/dev/audit-debug-log-retention.md](docs/dev/audit-debug-log-retention.md) | Audit debug log retention policy + runbook: sweep locations/thresholds, dry-run/`--apply`/`--check`, cron install/removal, safety guarantees |
 | [docs/dev/downtime-dispatcher.md](docs/dev/downtime-dispatcher.md) | Leader-election + shared-coordination downtime dispatcher architecture (herdr); critical-first dispatch tier + freeze split-by-skill; dedicated Dispatcher workspace with per-project tabs (one tab per work-item prefix) |
+| [docs/pane-triage-skill.md](docs/pane-triage-skill.md) | Approval-gated pane-triage skill and the cross-language pane-close bridge that exposes the shared classifier to Python skills |
 
 ## Tutorials
 

docs/design/mode-switch-worker.md

@@ -43,6 +43,26 @@ presence and proxy idle state, without operator intervention:
   responses yield no switch and never crash or block the plugin. A `409`
   (mode-switch restart in progress) is a no-op, retried on a later tick.
 
+### Drain before switching (WL-0MUL0KO7Q003O7YJ)
+
+When the idle window is met **and** the proxy is idle, the worker enters a
+**draining** state instead of switching immediately: it waits until the active
+session count has fallen to the cheap pool's budget (`freeSlots >=
+DOWNTIME_AUDIT_MIN_FREE_SLOTS = 2`), then switches. In-flight work is never
+killed — draining only delays the switch and blocks *new* dispatches.
+
+While draining:
+
+- The downtime dispatcher pauses **new** dispatches. The mode-switch worker
+  exposes `getIsDraining()`; `index.ts` forwards it to the dispatcher as
+  `config().drainPaused`, and `DowntimeWorker.tick()` refuses to dispatch while
+  it is `true`. The poll and `onProxyIdle` hook still run every tick, so the
+  mode-switch worker keeps observing the live free-slot count.
+- Once the budget frees, the drain completes, the proxy switches to cheap,
+  `getIsDraining()` clears, and dispatch resumes on the next idle tick.
+- An operator agent-route command (`onOperatorCommand`) cancels the drain
+  immediately, resets the idle clock, and keeps the proxy fast.
+
 ### Restart semantics
 
 On plugin/pane restart the idle clock resets to "active now" (construction

docs/dev/context-budget.thresholds.json

@@ -1,6 +1,6 @@
 {
   "global_agents": 0,
   "project_agents": 1655,
-  "skills_prose": 239,
-  "total": 1894
+  "skills_prose": 473,
+  "total": 2128
 }