Skip to content

kane-cli 0.8.2

Choose a tag to compare

View all tags
@robot-lt robot-lt released this 14 Aug 08:41
· 14 commits to main since this release
0.8.2
c0ecdd0

Highlights

The 0.8.2 release overhauls how test coverage and review results are displayed — cleaner tables, accurate status bars, and a new three-line AssuranceBar replace the old layout. Mobile app workflows now call the API directly, session management gains device persistence and prepared-device reuse, and a wave of display and export fixes clean up column math, word-wrap, and replay fidelity.

[0.8.2] - 2026-08-13

Mobile workflows simplified

  • App upload, list, and download go direct — these commands now call the API directly instead of going through the extra service hop.
  • doctor --target is now mandatory and singular — the flag requires exactly one target; dependency status matches on bundle identity for accurate results.
  • Mobile test creation uses the correct route — mobile test create now shows the correct Mobile category on test manager.

Session and replay reliability

  • Device persistence and prepared-device reuse — sessions now lock to their target, persist device selection across reconnects, and reuse prepared devices when available.
  • Scroll recording is replay-safe — scroll-into-view actions are recorded as SCROLL_UNTIL_ELEMENT and written to the replay/export record so they survive round-trips.
  • JSON path in sub-check results is preserved — json_path is now stored, so replay and export honor the original check structure.
  • Evidence steps show human-readable captions — step labels in evidence output use descriptive text instead of raw locators or bare "Step N" strings.
  • Pause, crash, and held summaries include credits used — these session-end states previously omitted credit counts; they are now reported correctly.

Sharper assurance views

  • Coverage table is high-level again — the main table lists only use cases and progress; detailed census data and commands have moved into the dossier, so the default output is scannable at a glance.
  • --rollup replaces --aspect — the flag was renamed to match existing JSON vocabulary (proven.aspect) is unchanged.
  • Gaps view is now the coverage ribbon — use-case bands show full-word progress bars on a shared grid; detail lives in the dossier.
  • New ContextBar — a compact, status-first three-line bar sits behind a full-width divider and shows aggregated per-turn state; the transcript keeps one line per turn instead of accumulating duplicates.

Review results you can trust

  • All-clear never claims proven without execution facts — the review round no longer declares a use case proven on percentage alone; it requires an actual execution count.
  • Single-UC JSON closes over its own use case — other entries are emptied so agent next-step suggestions can't leak commands from unrelated project use cases.
  • Dossier acceptance criteria sort numerically — AC-2 now appears before AC-10 instead of sorting lexicographically.
  • ID columns cap at 16 characters — long slugs no longer push column widths negative or break the layout.
  • Action commands in the dossier wrap at the label column — long commands hang-wrap cleanly and never break mid-word at the terminal edge.
  • Review fold IDs never clip — the id column is now dynamic; class columns fit every word; degraded classes keep their commands; --stage no longer reorders bands.

Install

# npm
npm install -g @testmuai/kane-cli@0.8.2

# Homebrew
brew install lambdatest/kane/kane-cli

# Shell installer (latest release)
curl -fsSL https://raw.githubusercontent.com/LambdaTest/kane-cli/main/install.sh | bash

Compare: 0.8.1...0.8.2

Assets 18
Loading