Repurposing Grok's app builder into a multi-agent 'literary machine' for research papers
kiltrout · reddit · 2026-09-08
Frustrated that LLMs hallucinate closure on long texts and produce weak prose, the author repurposed Grok's app builder into a "literary machine": a multi-agent pipeline seeded with axioms like "problems are open vectors to be closed by research, not invention."
- Fetcher/cleric agents build a weighted corpus of sources under domain-tagged "expert" agents
- A Synthesist agent ("Inspiraation") provides panoramic overview, draws the thesis, and gives data a narrative shape
- A "breath" agent rewrites output against Wikipedia's list of AI-writing tells to sound human
The result was a clear improvement in coherence and sophistication over single LLM prompts — at enormous cost: a week's worth of Grok Pro credits and many hours of corpus building. Early outputs were merely formal data lists; the synthesist stage is what made them read as essays. Remaining issues include context bleed and voice tuning. Agent names carry deliberate misspellings to detect context bleed.
More from coding & agent
- SF 'Code Engineering' event picks up where 1,000-person harness engineering meetup left off — LangChain · 2026-10-27
- Prompt turns Claude Code into an iPhone backup forensics tool exposing 30-day 'deleted' photo copies — alex_verem · 2026-10-27
- Prompt turns Claude Code into a forensics tool finding deleted iPhone photos in backups — alex_verem · 2026-10-27
- Phones won't be mobile IDEs but control planes for agent fleets, argues Yuchen Jin — Yuchenj_UW · 2026-10-27
- FrontierCS 2 launches on Kaggle: 48 tasks make agents beat published CS papers — MeganRisdal · 2026-10-27
- Agent Field Kit launches $29 security runbooks for production AI agents — alvelda · 2026-10-27