Open
Conversation
Add sequential script execution to Test screen for testing persistent browser sessions. Users can now run two scripts back-to-back with the browser kept alive between runs. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
…er.py The persistent browser session feature passes --server-mode to whichever Python wrapper is selected. On Windows with browser_engine=computer_agent, this routes to computer_agent_wrapper.py which was missing the flag. Delegates server mode to OpenAIPlaywrightExecutor.run_server_mode(). Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Two fixes for persistent browser session on Windows: 1. Rust: Compact JSON to single line before writing to stdin. Pretty-printed JSON with newlines breaks Python's readline()-based NDJSON protocol. 2. Python: Set UTF-8 encoding on stdin/stdout in server mode on Windows. Default cp1252 encoding causes OSError on pipe writes. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
…erver mode (v0.5.2) - Add select_by_llm action: extracts all <select> options via DOM, sends to text-only LLM for matching, uses select_option() to fire native change event (fixes Angular submit button staying disabled) - Harden server mode: wrap execute_script in try/except to prevent unhandled errors from killing the persistent Python process - Add browser health check with auto-recovery in server mode loop - Use json.dumps(default=str) to handle non-serializable result objects - Update address selection strategies: regex (fast) -> select_by_llm (reliable) -> vision LLM (fallback) - Bump version to 0.5.2 Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
…e extraction (v0.5.4) Replace CSS-selector-based extract_dom (72% failure rate) with extract_by_llm that sends cleaned page HTML to the LLM for field extraction. Strips non-content elements, truncates to 50K chars, and uses JSON mode for structured output. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
…(v0.5.5) - Add Windows UTF-8 stderr encoding to prevent crashes from Unicode characters - Fix escalate strategies silently succeeding without doing anything - now fail explicitly with NotImplementedError - Add validation for empty strategy steps - Truncate long error messages (Playwright timeouts) to 500 chars to prevent I/O issues - Replace broken escalate fallback with working wait-and-retry select_by_llm The escalate feature was never implemented in workflow_executor but strategies with escalate were marking themselves as successful, causing workflows to continue with unselected addresses and eventually crash. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.