Repository navigation
v2026.11.09
Traces is now the online view of a PLC. Press Traces and every NET shows what Home Assistant is doing right now: the state of each block, the two values a comparison compares, a TON counting down, and what the last real run did, written under each block.
FORCE TRUE runs the trigger you choose. In Real, click a trigger and press FORCE TRUE ▶ run it in Home Assistant. Home Assistant runs the saved automation for real, as if that trigger happened now, so the rung of that trigger runs. Before, a run started with no trigger at all, from LIBI and from Home Assistant's own Run, so a rung that runs for its own trigger never ran. In Simulate, the default, nothing is sent to Home Assistant: FORCE a block TRUE or FALSE and watch the NET run inside LIBI.
Nothing can be edited while Traces is on. A click only follows the run or FORCEs a block. Turn Traces off to edit.
This release also fixes Traces showing one NET with another NET's states. With two automations opened from Home Assistant, a lamp could be shown off right after it was turned on, and a rung of the other NET lit.
Before you update
- Restart Home Assistant after the download. HACS shows Restart required.
- Reload LIBI's page in every browser and in the Home Assistant app. A LIBI page left open from before the update saves nothing and says "LIBI was updated: reload this page".
- If LIBI says it could not draw this page, reload without the cache: Ctrl+Shift+R, or Cmd+Shift+R on a Mac.
- In the Home Assistant app, reset the frontend cache in the app's settings.
Nothing else is required. Your projects, NETs and version history open as before, Traces never writes to your automations, and the minimum Home Assistant version is still 2025.1.0.
What Traces shows
- Open it: press Traces in the header (on a phone, in the menu). The ONLINE bar opens over the NETs, with the clock and the switch between Simulate and Real.
- Over each NET, a monitor shows:
- the clock now
- whether its automation is on or off in Home Assistant, and whether it is running
- its last run
- the value of every variable it uses; one that is offline is red
- Each block shows:
- its state: green when it is TRUE, or for an action while what it drives is on, and grey when it is not
- OFFLINE or NOT IN HA when its entity is unavailable or gone
- for a comparison, IN1 and IN2, for example the clock now and the time it waits for
- for a CMP DAY, the day now and the days ticked on the card, by name: IN1 sun, IN2 sun mon tue wed thu. The monitor shows the days by name too.
- for a comparison with a list of states, such as IN2 rainy, pouring, hail: = is TRUE when IN1 is one of them, and ≠ when it is none of them, as Home Assistant checks it. While it rains, a watering rung's ≠ rain is FALSE, and a simulated run stops there.
- for a comparison with a zone, IN1 against the zone's name, as a person's state names it
- for a TON, the time left, counting down: the for of a trigger, or the delay a run is in
- The last real run, from Home Assistant's own trace, is written under each block:
- RUN ▶ started the run: the trigger that started it
- RUN ✓ true and RUN ✓ done
- RUN ✗ false: stopped here
- RUN ⚠ error, with Home Assistant's reason in the block's menu
- RUN ⏳ 00:01:20: the step it is in, counting down
- RUN – not reached
- A run started with no trigger says so in the monitor. For a NET you drew in LIBI, the run is shown in its monitor rather than on each block.
- NETs are taller while Traces is on, so that every line under a block has room. Turn Traces off and they are drawn as you edit them.
FORCE: click a block
- Click any block to open its menu, next to the block. It shows what the block gives now, what the run did there, and what you can do with it.
- Simulate (the default):
- FORCE TRUE or FORCE FALSE. Release frees the block, and Release all in the ONLINE bar frees every block.
- A trigger forced TRUE runs its rung inside LIBI only: each condition, each action it would run, each TON counting down (⏭ ends one at once). The lines under the blocks start with SIM.
- Nothing is sent to Home Assistant, and no device moves.
- Real:
- Click a trigger, then FORCE TRUE ▶ run it in Home Assistant. LIBI names the automation, the trigger and what it switches, and asks first. When the automation has conditions of its own, you can also run it without them.
- Home Assistant runs the saved automation as if that trigger happened now. The result appears under each block. In Home Assistant, the run's trace says LIBI: FORCE TRUE.
- Home Assistant checks the conditions of each rung itself. To force a condition, use Simulate: a condition's menu has Switch to Simulate.
- What the trigger itself would tell (
trigger.to_stateand the like) is empty in such a run. - An automation whose mode is single and that is already running ignores the run, as Home Assistant does.
- Run without a trigger, over each NET, is Home Assistant's own Run: a rung that runs only for its own trigger does not run.
What Real does not run
- An automation that can reach a lock, an alarm or a cover. That covers one named in the automation, or in any script or automation it runs, however deep. It also covers one in an area, floor, device or label it targets, and one in a scene or a group it turns on. Try it in Simulate, or run it from Home Assistant's own editor.
- An automation LIBI cannot read through: one that picks what it switches with a template, or that calls a script Home Assistant has not loaded.
- A NET with changes not saved. Home Assistant would run what it has, not your drawing. Nor a NET not in Home Assistant yet, or an automation that is turned off.
- An automation changed after you were asked. LIBI checks it again just before the run. If it changed in Home Assistant in the meantime, LIBI does not run it and says so.
- Not checked: another automation that starts because of what the run switched. It runs as it would if you had switched that yourself.
Read only while Traces is on
- A click only follows the run or FORCEs a block. Everything below waits until Traces is off:
- choosing or changing an element, the Inspector, and the entity chooser
- adding, moving and deleting: the + points, Add Rung, dragging, dropping a variable, and right-click Rename
- a NET's name and kind, Load, Versions, Duplicate and Delete NET
- Undo and Redo (and Ctrl+Z, Ctrl+Y), SAVE & COMPILE, and Clear all nets
- opening or creating a project, the AI Chat, new variables and helpers, and Global
- A click on one of those says "Traces is on, so nothing can be edited. Turn Traces off to edit."
- Still there: folding a NET, its YAML and its dashboard card, the Variables bar's tabs and search, Logs, and the automation's on/off switch.
Fixes
- Traces drew a NET with another NET's states. Two automations opened from Home Assistant got the same block ids, so Traces drew one with the other's states: a lamp shown off right after it was turned on, and a rung of the other NET lit. Each NET now has its own.
Under the hood
- New websocket command
libi/online_fire, admin only. It writes no file.- Without
fire, it only says what a run would be, or why LIBI does not run it. - With
fire, sent only after the user confirms, it makes the plan again. It runs the automation only when it is still the one the user was shown (before).
- Without
- The run is the automation entity's own
async_trigger, which Home Assistant'sautomation.triggercalls too. It passes the trigger'sidandidx. Home Assistant checks the conditions, honours the automation's mode and keeps the trace. - What a run can reach is read from what Home Assistant has loaded, never from LIBI's drawing.
libi/project_openandlibi/save_donealso returnkeys, each NET's key in Home Assistant. Traces uses it to find the automation and the trace of a NET drawn in LIBI.- Panel build 3.94.3.
- Tests: 382 pytest tests against a real Home Assistant, and 40 browser test pages. Traces is tested on runs captured from a real Home Assistant.
Full Changelog: v2026.11.06...v2026.11.09
What I checked:
GitHub, on 8762542: CI, Hassfest and Validate (the HACS Action) all passed.
Gate with GITHUB_REF_NAME=v2026.11.09: BLOCKER 0, ERROR 0, HOT_WARN 0, WARN 1, INFO 2, the same as before.
libi.zip built from the current code: 31 files, only the integration, no pycache.
Read-only review of what changed since 2026.11.08: only drawing in Traces, versions and tests. Nothing calls Home Assistant, saves, deletes or writes HTML.
pytest (382) and the panel pages (40) ran here on the same code, before the version bump. CI also ran pytest on GitHub after the bump, and it passed.
One open issue, your call (compile #11):
Traces now compares against a list of states the way Home Assistant does.
But a CMP with a list that you create or edit in LIBI is still saved as a single text value. So for such a NET, Traces can show FALSE while Home Assistant treats the condition as TRUE.
The same bug exists in v2026.11.06.
Automations imported and left unedited are saved as they were, so they're fine.
Full Changelog: v2026.11.06...v2026.11.09