v0.2.2

Ordinus 0.2.2

Token usage tracking

Every run now records how many tokens it used: input, output, cached, and the run's real cost computed from the provider's raw counters. Codex reports cumulative session totals, Claude and Gemini report per-run totals; both are handled correctly. The data is stored per run and will feed the run inspector later.

Lower token consumption

• Resumed runs no longer repeat the workspace and output-format rules on every message. The session already has them; a one-line reminder is enough.
• When an agent continues its own work in the same session, its previous output is no longer sent back to it. It only receives a summary and a run id.
• These changes also improve provider cache hit rates.

Follow-up work can now see earlier results

Until now, a follow-up item added to an existing Work Request could not read the output of earlier items unless you linked them by hand. Three things fix that:

• The planner can now bind new items to existing completed runs on its own.
• The plan review dialog lists the request's earlier items, shows what the planner linked, and lets you change the links before starting.
• Each Work Request keeps a digest.md file in its working folder: a plain record of what was done, by whom, with run ids. Agents can fetch the full output of any listed run through a new read-only tool.

Better plans

The planning prompt was rebuilt. It now sizes plans from the actual boundaries in the request instead of leaning toward fewer items, writes its reasoning before producing the plan, checks itself before answering, and treats your agent selections as real choices instead of hints. A new example shows a healthy multi-agent plan.

Agent creation fix

When you described a new agent, the text you typed was saved as the agent's capabilities, overwriting the proper capabilities line the AI had generated. This misled the planner when assigning work. Fixed; the generated line is kept. New agents also get a short title as their role instead of a long sentence.

Security

The internal tool server now uses token-gated endpoints. Worker agents can only reach their own request's read-only tools and can no longer reach the assistant's full tool catalog by changing the URL.

Small fixes

• The @agent suggestion panel in the new request dialog now opens at the cursor instead of the bottom of the box, and the list scrolls along when you navigate with arrow keys. The same scroll fix applies in Conversations.
• Clicking an agent's name on a Work Item now opens the Agents page with that agent selected.
• Plan links to failed or cancelled runs are no longer offered, since they would block the new item forever.

v0.2.1

v0.2.1

Workboard

• The run detail modal now shows the original brief ("What you asked for":
  instruction and expected output). For finished work it sits collapsed below
  the agent's message; for running, cancelled, or queued work it is shown
  expanded, since there is no output yet to read.
• Replaced the narrow right-side inspect panel with a full-width bottom sheet.
  It has a compact metadata line (agent, provider/model, sandbox, start time,
  session id) and two tabs: Activity, a row-based timeline of what the agent
  did, and Console, the sanitized invocation plus raw stdout/stderr with
  proper line wrapping, auto-follow, and copy.
• New ways to open the inspector: a terminal icon in the modal header, and
  clicking the progress box while a run is in flight.
• The progress box now uses the same rotating border animation as the rest of
  the app instead of the misplaced orange bar.

Home and agent rooms

• The same inspector is now available in Home and in agent rooms. Clicking
  the live status line opens it for the turn currently running; hovering a
  finished reply reveals an icon that opens the record for that turn. Records
  are persisted, so this also works after a restart.
• Added a small read-only IPC method (observability.getTurnRun) to look up
  the observed run behind a single conversation turn.

Earlier in this release

• Left rail design system: a shared borderless record list across screens,
  with the workflows screen refactored to use it.
• Live activity line: the static "thinking" label was replaced by a single
  mutating status line fed by the observability pipeline, with a Stop button
  while a turn is running.
• Transcripts now list the files a turn created or changed.

Fixes

• Clipboard copy buttons in the inspector no longer fail silently; copying
  falls back to a second method when the Clipboard API is unavailable.
• Long console output is readable: text wraps instead of being cut off at the
  panel edge.

v0.2.0

Ordinus 0.2.0
Getting started is now much smoother — a guided first run that installs the AI CLIs for you — and the multi‑agent side got real depth: agents that discuss decisions, durable hand‑offs, and folder‑scoped isolation.

New features

• Guided first‑run onboarding — on first launch Ordinus walks you through setup and installs and wires up the AI CLIs (Claude, Codex, Gemini) for you. No terminal, no manual setup.
• Home — a new front door: ask Ordinus directly from a Home panel, with a welcome flow for new users.
• Advisory Conversations — multi‑agent discussions where a moderator routes a topic to several agents that actually hear each other, build on or challenge one another, and end with a synthesis — bounded and controllable.
• Folder‑scoped agents & workspace folders — scope agents to specific folders so their work stays isolated and organized.

Improvements

• Durable results & hand‑offs — work results are now database‑backed, so outputs and agent‑to‑agent hand‑offs persist reliably.
• Better run output — smarter summaries with size limits and robust extraction of fenced JSON from agent output.
• Internal refactor for readability and maintainability.

⚠️ Note
Builds are unsigned — macOS may show a "damaged"/Gatekeeper warning and Windows may show SmartScreen on first launch. See the README for the one‑time workaround.

v0.1.0

Full Changelog: https://github.com/muratgur/ordinus/commits/v0.1.0