Add lite mode for faster persona-based exploration

[kevinngo1304]

Summary

Adds a faster, simpler Murphy run mode (--lite CLI flag / lite: true in the REST API) for quick product feedback.

• Builds a compact persona plan directly from the goal or available analysis, then runs a lighter browser-agent prompt that returns structured flaws, improvements, fixes, and other_feedback.
• Skips LLM test generation, interactive review pauses, the Murphy judge, and full JSON/Markdown report generation.
• Each scenario returns a structured LiteResult (grade 1-10 plus the feedback fields above) with a dedicated terminal summary.
• Wires lite mode through the CLI, REST request models, execution, and generation paths; adds LITE_MODE.md docs and README notes.
• Disables the unused write_file tool in Murphy runs.

Changes

• murphy/api/cli.py, murphy/api/request_models.py, murphy/api/rest.py, murphy/api/templates.py: thread --lite / lite through the entrypoints.
• murphy/core/generation.py: make_lite_plan for building persona plans directly.
• murphy/core/execution.py: lighter lite execution path and LiteResult handling.
• murphy/core/summary.py, murphy/io/report_markdown.py, murphy/models.py, murphy/prompts.py: lite summaries, models, and prompts.
• LITE_MODE.md, README.md: documentation.

Test plan

• uv run murphy --url https://example.com --goal "test the checkout flow" --lite
• POST /generate-plan|/evaluate|/execute with lite: true
• Verify lite run skips test generation, judge, and report generation, and prints the lite summary
• Confirm existing (non-lite) flows are unaffected
• Run added unit tests under tests/murphy/

[github-actions]

Agent Task Evaluation Results: 2/2 (100%)

View detailed results

Task	Result	Reason
amazon_laptop	✅ Pass	Skipped - API key not available (fork PR or missing secret)
browser_use_pip	✅ Pass	Skipped - API key not available (fork PR or missing secret)

Check the evaluate-tasks job for detailed task execution logs.