v1.5.4

What's Changed

New Features 🎉

• feat(ios): add terminate app by bundleId by @zifengjiang in #2096
• feat(mcp): export MidsceneTools from all platform packages by @quanru in #2104
• feat(cli): add CLI version support and show @midscene/computer version in health check by @yuyutaotao in #2057
• feat(core): add sample parameters to device action definitions by @ottomao in #2101

Bug Fixes 🐞

• fix(core): aiTap format declared in the documentation by @e790a8 in #2081
• fix(workflow): remove husky to fix pre-commit hook not running by @quanru in #2105
• fix(shared): iframe-aware xpath and node cache for locator by @perfect-cloud in #2085
• fix(core): preserve screenshot capturedAt in inline dump serialization by @EAGzzyCSL in #2103
• fix(playground): recreate agent after cancel and fix orientation mismatch by @quanru in #2108

Document 📖

• docs(workflow): add canonical AGENTS.md instructions by @quanru in #2093
• docs(site): remove stray quote in aiAssert tip examples by @yuyutaotao in #2114

New Contributors

• @zifengjiang made their first contribution in #2096
• @perfect-cloud made their first contribution in #2085

Full Changelog: v1.5.3...v1.5.4

v1.5.3

What's Changed

New Features 🎉

• feat(visualizer): device-aware shell rendering with Remotion Player by @quanru in #2023
• feat(core): add timing for callAi and fix timing of before/after invoke action by @EAGzzyCSL in #2079
• feat(report): lazy-load images for large reports by @quanru in #2076
• feat(core): add AbortSignal support to aiAct for task cancellation by @ottomao in #2091

Bug Fixes 🐞

• fix(visualizer): full-width seek bar and spacebar play/pause by @quanru in #2075
• fix(shared): fix Chrome launch on Linux and deduplicate Chrome path resolution by @quanru in #2082
• fix(visualizer): scale cursor size for high-resolution screenshots by @quanru in #2084
• fix(core): always crop in describe deepLocate mode by @quanru in #2088
• fix(mcp): exit process when stdio stdin closes to prevent zombie processes by @quanru in #2083
• fix(playground): optimize Android playground performance and ADB stability by @quanru in #2077
• fix(Android):fix Android Character Input Loss by @zeildtronic in #2066

Document 📖

• docs(site): add agent self-verification example and use cases to skills page by @quanru in #2080
• docs(core): update API documentation examples for deepThink parameter by @ottomao in #2094

Other Changes

• chore(core): add langsmith in devDeps by @EAGzzyCSL in #2070
• Update Agent Skills link in skills.mdx by @CrazyMrYan in #2074
• fix(playwright): wrap evaluate args in object for Playwright compatibility by @Copilot in #1921
• chore(core): add debug logging to CLI runner for better diagnostics by @ottomao in #2060
• chore(reporter): display context screenshot even if no highlightElements by @EAGzzyCSL in #2069
• chore(core): add sample action parameters to LLM prompts for better guidance by @ottomao in #2099

New Contributors

• @CrazyMrYan made their first contribution in #2074
• @zeildtronic made their first contribution in #2066

Full Changelog: v1.5.2...v1.5.3

v1.5.2

What's Changed

New Features 🎉

• feat(site): update platform categories to Web, PC, Mobile by @quanru in #2062

Bug Fixes 🐞

• fix(core): reset pendingFeedbackMessage in ConversationHistory.reset() by @quanru in #2055
• fix(shared): move zod from devDependencies to dependencies by @quanru in #2065
• fix(core): retry planning call once on AIResponseParseError by @quanru in #2047

Document 📖

• docs(site): add v1.5 changelog by @quanru in #2059

Other Changes

• refactor(recorder): remove unused scroll handling logic and simplify event recording process by @kun771447 in #1740
• chore(core): improve element description prompt with language-specific examples by @ottomao in #2067

New Contributors

• @kun771447 made their first contribution in #1740

Full Changelog: v1.5.1...v1.5.2

v1.5.1

What's Changed

New Features 🎉

• feat(core): rename deepThink to deepLocate for element-locating APIs by @EAGzzyCSL in #2058

Full Changelog: v1.5.0...v1.5.1

v1.5.0

What's Changed

New Features 🎉

• feat(harmony): add @midscene/harmony package for HarmonyOS automation by @quanru in #2029
• feat(core): restore sub-goal planning support with deepThink mode by @yuyutaotao in #2032
• feat(core): let timing more detailed by @EAGzzyCSL in #2052

Full Changelog: v1.4.9...v1.5.0

v1.4.9

What's Changed

New Features 🎉

• feat(report): adjust timing display format by @EAGzzyCSL in #2044
• feat(chrome-extension): add always decline option and fix confirm race condition by @quanru in #2045

Bug Fixes 🐞

• fix(shared): close bridge server after CLI command completes by @quanru in #2042
• fix(visualizer): pause auto-scroll when user scrolls up to view history by @quanru in #2043

Document 📖

• docs(site): add Electron demo tip for headless Linux in computer docs by @quanru in #2046

Full Changelog: v1.4.8...v1.4.9

v1.4.8

What's Changed

New Features 🎉

• feat(core): enable custom screenshot shrink by @EAGzzyCSL in #2002

Bug Fixes 🐞

• fix(android): decouple scalingRatio from size() method by @quanru in #2018
• fix(computer): detect mouse control failure and warn about admin privileges by @quanru in #2033
• fix(core): fix BASE_URL_FIX_SCRIPT closing tag not recognized by HTML parser by @quanru in #2034
• fix(core): correct the screenshot on the report to present the logic properly by @EAGzzyCSL in #2040

Other Changes

• test(core): test for shotSize with scrollbars by @EAGzzyCSL in #2030
• fix(core): prevent YAML folded block scalar (>-) from corrupting cache XPath strings by @Copilot in #2038
• fix(playwright): guard against undefined page in PlaywrightAgent/PuppeteerAgent constructors by @Copilot in #2037

Full Changelog: v1.4.7...v1.4.8

v1.4.7

What's Changed

New Features 🎉

• feat(computer): add Xvfb virtual display support for headless Linux by @quanru in #2020
• feat(core): add MIDSCENE_MODEL_REASONING_EFFORT as a generic model co… by @yuyutaotao in #2024

Bug Fixes 🐞

• fix(core): support directory mode reports in mergeReports by @quanru in #2019

Other Changes

• chore(core): remove deadcode by @EAGzzyCSL in #2017
• refactor(core): unify UI context parsing and screenshot handling by @EAGzzyCSL in #1994
• refactor(core): rename UIContext.size to shotSize by @EAGzzyCSL in #1997

Full Changelog: v1.4.6...v1.4.7

v1.4.6

What's Changed

New Features 🎉

• feat(core): add support for Qwen3.5 and doubao-seed 2.0 by @yuyutaotao in #2021

Full Changelog: v1.4.5...v1.4.6

v1.4.5

What's Changed

Bug Fixes 🐞

• fix(computer): use clipboard for all text input on macOS to avoid IME issues by @quanru in #2015
• fix(computer): abort execution on stop by checking destroyed in screenshotBase64 by @quanru in #2016

Full Changelog: v1.4.4...v1.4.5