npm
Aspect Code
Open-source CLI that generates and maintains repository context for AI coding assistants.
Aspect Research · Independent lab
Studying how reliable AI systems behave in practice.
An independent lab studying how AI systems behave in practice. We ship tools, publish research, and run exploratory experiments.
01
Tools & systemsProjects
early stage: v0.X.X
Aedos
Hypothesize-and-verify architecture for reducing LLM hallucination. Extracts claims, routes them to verifiers, and accumulates a verified cache over time.
02
Methodology & writingPapers
PreprintJan–Apr 2026
Iterative Probe-and-Refine Tuning of Repository Guidance for AI Coding Agents
A methodology for refining repo-level guidance via synthetic behavioral probes; ~50% relative improvement on SWE-bench Verified resolve rate over 4 trials, and evidence of cross-model degradation.
Read →PreprintApr 2026
Structural Grammar of the Voynich Manuscript: Characterization and Mechanism Discrimination
Run by an autonomous multi-agent AI loop: roughly 600 corpus-statistical experiments and 5,400 cipher-mechanism simulations, with independent meta-review at phase boundaries.
Read →Forthcoming2026
Hypothesize-and-Verify with a Tiered World-Model Cache
Companion methodology to Aedos: how an assistant can extract claims, route them to type-matched verifiers, and accumulate ground truth over time without forgetting.
Join the Slack →03
People
04
Mission Statement
Reliable AI is a research problem. The lab investigates how AI systems behave in practice, builds the tools needed to study them, and publishes what it learns.