Agent wrote 41,000 lines over 11 hours to "protect" a simple DB migration
JeremyNguyenPhD · x · 2026-08-27
Rob Lennon shares his Sol workflow: use high-effort modes for planning and high-stakes work, cruise on medium the rest of the day — the model is a "striver" that earnestly wants to do well.
He also warns your agents.md may signal you expect rigor everywhere: in a recent /goal session, an agent wrote 41,000 lines of code over 11 hours to protect a simple database migration, essentially building a whole software suite with tests for one small transformation.
Related event: Over-testing AI coding agents drive up token costs(3 posts)→
More from coding & agent
- New Tool Runs Codex, ClaudeCode, OpenCode, or Pi Side-by-Side — tom_doerr · 2026-08-27
- DocETL: New Python Library for Agentic Data Processing and ETL with AI — mdancho84 · 2026-08-27
- Beyond the model: what's in your agent setup? — External-Wind-5273 · 2026-08-27
- Codex users would trade 5% quota to let running tasks finish — vinmi · 2026-08-27
- qmax-mcp: local QA MCP server that scans URLs and runs Playwright repros for coding agents — bestofdesp · 2026-08-27
- MCP client SDK for Apple's on-device model dynamically builds Tools from JSON Schema — AdventurousKeys · 2026-08-27