LLM Tails cuts agent input tokens 80% by making agents read before they write
savadks1818 · reddit · 2026-10-27
A developer released LLMTails, an open-source (MIT, zero-dependency) prompt + hooks + skills bundle built on the Ponytail and LLM Wiki ideas. It forces agents to read before writing, write minimal code, verify before claiming, and match your codebase; it works with Claude Code, Cursor, Copilot, Codex, Gemini, Windsurf, Cline and any agent that reads files.
Same-prompt A/B test building a small game (with vs without):
- Total input tokens: 1,009,551 vs 5,052,587 (-80%)
- Output tokens: 20,790 vs 74,311 (-72%)
- API calls: 11 vs 38 (-71%)
- Game code: 347 lines vs 1,041 lines (-67%), all tests passing in both
- Screenshots viewed: 1 vs 18
Net effect: far fewer tokens and hallucinations, with a leaner codebase.
More from coding & agent
- Precompute golden answers and match at runtime: an LLM pattern from AustinLangChain — colinmcnamara · 2026-10-27
- Dev argues there's still alpha in working with your coding agent interactively — alex_frantic · 2026-10-27
- Researchers launch AMA on patterns for safer AI agents used by citizen developers — Loose_Rip359 · 2026-10-27
- Gradio Adds In-Chat Human-in-the-Loop So ML Intern Pauses Before Training — Gradio · 2026-10-27
- Building Halo Infinite Forge maps with local LLMs, DCjson and Blender MCP — swagonflyyyy · 2026-10-27
- Ex-Unity dev Lucas Meijer says OpenClaw looks cleaned up and worth another run — lucasmeijer · 2026-10-27