v0.2.3

What's Changed

• discouraging use of screenshots for verification by @angiejones in #66

Full Changelog: v0.2.2...v0.2.3

v0.2.2

• Add server instructions to guide agents toward using accessibility://current resource
• Update take_screenshot tool description to prefer accessibility snapshot over screenshots
• Fix McpServer constructor to pass instructions in ServerOptions (second arg) instead of serverInfo

v0.2.1

What's Changed

• feat: accessibility tree by @angiejones in #64

Full Changelog: v0.2.0...v0.2.1

v0.2.0

What's Changed

• Add AGENTS.md for AI agent onboarding and MCP spec compliance by @angiejones in #47
• fix: use By.tagName() for tag locator strategy by @angiejones in #48
• test: add regression test suite by @angiejones in #49
• docs: update AGENTS.md — remove outdated info, add testing philosophy by @angiejones in #50
• fix: MCP spec compliance — isError, mimeType, image content type by @angiejones in #51
• Add CI workflow for automated testing by @angiejones in #52
• add MCP server badge by @punkpeye in #46
• feat: add new tools for windows, alerts, frames by @angiejones in #54
• fix: improve switch_to_frame API, add missing docs/tests, fix test reliability by @angiejones in #56
• feat: add Safari browser support by @angiejones in #57
• feat: add cookie management tools by @angiejones in #59
• Add WebDriver BiDi support for real-time browser diagnostics by @angiejones in #58
• Refactor: consolidate 32 tools → 18 by @angiejones in #61

Full Changelog: https://github.com/angiejones/mcp-selenium/commits/v0.2.0