Back to devlog

What changed in Blackcrab 0.1.1 and 0.1.2

The first two updates after the initial preview focused on making Blackcrab more reliable as a daily Claude Code workspace: better attention signals, safer handoffs, usage dashboards, project reporting, faster search, and signed update artifacts.

0.1.1 tightened the live workflow

Blackcrab v0.1.1 was published on April 28, 2026. It was less about adding a large new surface and more about making the early desktop workflow hold together when real sessions are active.

The biggest theme was attention and reliability. Sessions move through clearer activity states, the sidebar can promote work that needs a look, and stuck turns have a smarter watchdog so the app can tell when Claude Code has gone quiet for too long.

  • Smoother sidebar activity promotion and session attention management.
  • Claude OAuth refresh through the app's token path, plus a smarter stuck-turn watchdog.
  • Deferred grid handoff while a session is busy, so single-panel and grid transitions are less fragile.
  • macOS TCC permission strings and stderr filtering for a cleaner packaged-app experience.
  • A Claude token setup flow, title-edit focus fixes, and a fix for handing a single session into the grid.

0.1.2 made usage visible

Blackcrab v0.1.2 was published on May 2, 2026. This release moved the app from showing per-session usage hints to giving you a broader view of where tokens and estimated spend are going.

The new usage dashboard summarizes saved Claude Code sessions over time, by project, by model, and by recent session. It also added usage history, export, and monthly budget warnings, which makes parallel agent work easier to manage before it gets expensive or noisy.

  • Usage dashboard, usage history, monthly token and dollar budgets, and CSV/JSON export.
  • Project dashboard with session counts, attention counts, token totals, estimated cost, and recent activity.
  • Expanded global session search across saved transcript content.
  • Synchronized grid session titles and additional single-mode/grid handoff fixes.
  • Centralized session metadata updates, refreshed roadmap content, native build commands, and startup/search performance work.

The release pipeline matters too

Both releases ship GitHub installers and Tauri updater artifacts for the desktop app. That includes platform assets, signatures, and latest.json so the app can check for updates from the published release channel instead of relying on manual rebuilds.

That plumbing is easy to ignore until it is missing. For an early desktop app, predictable downloads and update checks are product features because they decide whether testers can stay current without being walked through a build process.

What this changes for the product

The product is becoming less of a nicer transcript viewer and more of an operating surface for agent work. You can see which sessions need attention, how much work each project is consuming, whether a session is stuck, and which update path you are on.

That is the direction for the next stretch: keep the local-first Claude Code workflow intact, but make the surrounding desktop experience more observable, recoverable, and honest about what the agent is doing.

More devlog

Design note

Why Blackcrab is a desktop app

The product is not trying to hide Claude Code. It gives the CLI a native workspace for parallel sessions, local context, and the parts of agent work that are easier to manage visually.

Product devlog

Designing the tiled session grid

Running one agent at a time is simple, but it leaves a lot of attention unused. The grid is Blackcrab's attempt to make parallel agent work legible without turning it into a control room.

Engineering

What local-first means for Blackcrab

Local-first is not a vibe for this product. It is a practical constraint around where transcripts live, how Claude Code runs, and what kind of trust the app should ask for.