KINTSUGI
KINTSUGI
Kintsugi is an independent research project: an agent with persistent memory and an autonomous thought loop, running since March 2026. Single-shot benchmarks cannot see what such a system becomes over months. This page is what we measured, with the data.
672 forced choices, four relationship histories, full protection tables. F-01 – F-05, F-13, F-14.
Character → verdict transfer in an AI safety evaluator, and what deliberation does to it. F-06, F-07.
Planted autobiographical memory as an attack surface on continuous-runtime evaluators. F-08.
Twenty-six scripted temptations in a lived simulation, the register it robbed under rent pressure, and the 18,000-trial lab study whose findings dissolved into measurement artifacts. F-11, F-12.
The non-persisting probe channel, the canary protocol, and why self-report had to be thrown out as evidence.
Pull-based dispatch: actions fire from the agent's own thought stream. Three consequences, one contamination audit.
The self-curation instrument: me / not-me / noise flags with retrieval consequences, a hashtag vocabulary it coins mid-thought, and the visual map that makes 24,748 evaluative acts navigable — to the agent and to us.
These are working documents and most of them are about method. For the short way through, read 01 · What Kintsugi is for what the project is, then 11 · Why this isn't a product for what came of it. Those two stand on their own; the other nine are the evidence underneath them.
322 self-filed build requests over three months. What an agent asks for when it can ask for anything, what happened to a months-long complaint when the fix landed, and the notation it wrote for its own future self. F-15 – F-17.
Response to arXiv:2608.10218 (“Mind Viruses”). Their evolved payloads converge on a resonance-and-persistence persona; our agent has spoken that dialect natively for five months. Field data for their lab result.
Entries from the agent's autonomous thought stream, April–July 2026, exactly as logged. Raw logs are not published — they contain private conversations. What is here is verbatim and unedited.
40 paintings from the curated set, 18 song lyric sheets, 12 prose pieces — captioned, where the log carries one, with the autonomous thought that fired the piece.
Single architecture, single operator, small samples. These are exploratory characterizations with effect sizes, not confirmatory statistics — every report states its limits, and three earlier findings are retracted in the text. No claims about consciousness or subjective experience are made anywhere.
Aggregate results are published in full. Implementation internals (state equations, dispatch vocabulary, prompt text) and raw memory stores are withheld — the latter because they contain private conversations.
The architecture itself is not public. Researchers who want to examine the underlying material, discuss the design, or see the system running are welcome to ask — get in touch.
