Build a web UI (Flask/FastAPI) to visualize past agent runs. Currently, run data is stored in JSONL and JSON files that must be manually inspected. A dashboard would show: run history with pass/fail status, per-iteration diffs and test output, LLM confidence trends across iterations, file change summaries, and cost/token usage per run.
Build a web UI (Flask/FastAPI) to visualize past agent runs. Currently, run data is stored in JSONL and JSON files that must be manually inspected. A dashboard would show: run history with pass/fail status, per-iteration diffs and test output, LLM confidence trends across iterations, file change summaries, and cost/token usage per run.