# Webwright | Microsoft | Knowledge Base | Instagit

A simple SWE style browser agent framework that achieves SOTA results on long horizon web tasks. 

GitHub Stars: 5.6k

Repository: https://github.com/microsoft/Webwright

---

## Articles

### [Common Webwright Errors and Format Failures: Troubleshooting Guide](/microsoft/Webwright/what-are-common-webwright-errors-and-format-failures-and-how-to-troubleshoot-them)

Troubleshoot common Webwright errors and format failures. Learn to resolve issues with nested loops, missing configs, invalid inputs, and malformed JSON output for the microsoft/Webwright repo.

- Tags: how-to-guide
- Published: 2026-06-25

### [Webwright Output Artifact Structure: Screenshots, Logs, and Trajectory Files Explained](/microsoft/Webwright/what-is-the-output-artifact-structure-in-webwright-including-screenshots-logs-and-trajectory-files)

Understand the Webwright output artifact structure. Learn how screenshots, logs, and trajectory files are organized for every run under your specified output folder.

- Tags: deep-dive
- Published: 2026-06-25

### [How to Use Persistent Browser Sessions with `persistent_local_browser` in Webwright](/microsoft/Webwright/how-to-use-persistent-browser-sessions-with-the-persistent_local_browser-tool-in-webwright)

Learn to use persistent browser sessions with Webwright's persistent_local_browser tool. Reconnect across scripts with a detached Chromium process and reusable CDP endpoint.

- Tags: how-to-guide
- Published: 2026-06-25

### [How to Integrate Webwright with OpenAI Codex as a Plugin: Complete Developer Guide](/microsoft/Webwright/how-to-integrate-webwright-with-openai-codex-as-a-plugin)

Learn how to integrate Webwright with OpenAI Codex as a plugin. This guide shows you how to configure the manifest to enable browser automation skills for Codex.

- Tags: how-to-guide
- Published: 2026-06-25

### [How to Build Repeatable Run Dashboards with Flask Using Webwright](/microsoft/Webwright/how-to-build-repeatable-run-dashboards-with-flask-using-webwright)

Build repeatable run dashboards with Flask and Webwright. Deploy the Flask app to view and re-trigger Webwright tasks from execution traces using the webwright run CLI.

- Tags: how-to-guide
- Published: 2026-06-25

### [How Webwright Handles History Compaction Using LLM Summarization After N Steps](/microsoft/Webwright/how-does-webwright-handle-history-compaction-using-llm-summarization-after-n-steps)

Learn how Webwright uses LLM summarization for history compaction. Discover how it bounds token usage by summarizing conversations every N steps with the system prompt.

- Tags: internals
- Published: 2026-06-25

### [How the Self‑Reflection Tool Validates Task Completion Using `predicted_label`](/microsoft/Webwright/how-does-the-self_reflection-tool-validate-task-completion-using-predicted_label)

Learn how the Webwright Self-Reflection tool validates task completion using the predicted_label signal. Discover how it determines success or failure for downstream agents.

- Tags: deep-dive
- Published: 2026-06-25

### [Webwright Workspace Structure Explained: plan.md, final_script.py, and finalruns/](/microsoft/Webwright/what-is-the-structure-of-the-webwright-workspace-including-plan-md-final-script-py-and-final-runs)

Understand the Webwright workspace structure including plan.md, final_script.py, and finalruns. Learn how Webwright organizes tasks and execution artifacts for efficient automation.

- Tags: architecture
- Published: 2026-06-25

### [How to Debug Webwright Runs Using the `--debug` Flag and Headed Browser Mode](/microsoft/Webwright/how-to-debug-webwright-runs-using-the-debug-flag-and-headed-browser-mode)

Debug Webwright runs with the --debug flag. Launch a headed browser with DevTools and inspect delays for effective troubleshooting.

- Tags: how-to-guide
- Published: 2026-06-25

### [How the Webwright Agent Loop Works: Message Flow and Action Execution Explained](/microsoft/Webwright/what-is-the-webwright-agent-loop-and-how-does-message-flow-and-action-execution-work)

Understand the Webwright agent loop: Explore the prompt observe execute cycle and how LLM agents control Playwright browsers through Python code. Learn message flow and action execution.

- Tags: internals
- Published: 2026-06-25

### [How Webwright's Terminal-Driven Architecture Differs from Traditional Browser-Agent Frameworks](/microsoft/Webwright/how-webwrights-terminal-driven-architecture-differs-from-traditional-browser-agent-frameworks)

Discover how Webwright's terminal-driven architecture revolutionizes browser automation by treating browsers as disposable, controlled environments unlike traditional persistent session frameworks.

- Tags: architecture
- Published: 2026-06-25