ChatWalaʻau v0.112.0

This release adds one major capability to ChatWalaʻau: Declarative Workflows -- YAML-defined orchestrations that run your declarative agents as a graph.

No breaking changes. Purely additive: with DECLARATIVE_AGENTS_DIR unset there are no workflows and everything behaves as before.

Highlights

Declarative Workflows (v0.112.0)

• Define a workflow in YAML (kind: Workflow) that orchestrates your declarative Prompt agents as an execution graph -- sequential steps, conditionals, loops, foreach, and agent invocations -- compiled from the file with the Microsoft Agent Framework. No new dependency.
• A workflow only references your agents by name; ChatWalaʻau builds each one itself, so credentials, provider, and sampling are never taken from the workflow file. An unsupported action or a reference to an agent that is not a Prompt agent is flagged and blocks the workflow from running, so a workflow that validates cleanly always runs.
• Workflows live in the same folder as your agents (DECLARATIVE_AGENTS_DIR), told apart by the kind field. A runaway-loop / cost guard (WORKFLOW_MAX_ITERATIONS, default 100) bounds every run.

One place to manage agents and workflows (v0.112.0)

• Prompt agents and workflows are now managed in one modal (the robot icon in the sidebar footer), told apart by a Prompt / Workflow tag.
• Activate a Prompt agent as before, or press Run in chat on a workflow to make it the chat run-target -- the choice is remembered. Create opens the matching editor (New Prompt agent or New Workflow).
• Every assistant message is labeled with the agent or workflow that produced it, so you can always see what answered.

Build a workflow as a diagram (v0.112.0)

• Create -> New Workflow opens a full-screen editor: a visual DAG canvas of your steps, a form for each step (an "Invoke agent" step picks one of your Prompt agents), and a live YAML preview you can also edit directly.
• Validation runs continuously and mirrors what running enforces, so warnings show while you edit.

Run it in chat or in the background (v0.112.0)

• In chat: your message drives the selected workflow and its progress is shown live as a small graph; each step's output is separated so the steps are easy to read. While a workflow is the run-target, the per-message model / reasoning / verbosity / structured-output controls are hidden, because each step's model is fixed by the agent it invokes.
• In the background: run the same workflow as a pipeline job with run history and a captured log, for long or unattended runs.
• While any turn is still running -- a Prompt agent or a workflow -- the assistant message keeps an animated "processing" indicator until it fully completes, so a workflow that pauses between steps no longer looks finished.

Included in this release

This release includes implementation work from:

• v0.112.0 through v0.112.0

Notes

• Internal design management content and architecture control artifacts are intentionally omitted from this release note.

Installation

See the repository README for setup and usage instructions.

Version

• Release version: 0.112.0
• Previous release: 0.111.0