ByeongkiJeong/MiroFish-Ko
Choose this fork if Korean localization is the main adoption constraint and you want a mostly upstream-compatible swarm-simulation platform. Choose upstream instead if you want the freshest fixes and the original multilingual presentation with less divergence risk.
Choose this fork if you want a more adoption-friendly, English-centered MiroFish with an extra backend/provider option and workflow-oriented changes. Stick to upstream if you want the cleanest baseline and the absolute latest upstream patch set.
Prefer this fork if you want MiroFish in an English-friendly package with some cleanup around simulation and reporting. Prefer upstream if you want the original documentation set and the lowest-divergence path for following future releases.
newway-anshul/MiroFish-EN
Prefer this fork if you want MiroFish in English and value added logging for LLM-driven runs. If you mainly want the newest upstream functionality with minimal divergence, upstream is still the safer default.
Prefer this fork if you want local Neo4j/Qdrant-backed memory, migration tooling, and more durable simulation state. Prefer upstream if you want the newest fixes and the broadest maintained feature set; this fork is more specialized and slightly behind.
Choose this fork only if you want a near-identical copy of MiroFish with minimal modification risk. If you want the newest upstream fixes, use upstream instead.
Choose this fork if you want a more localized, operationally convenient MiroFish experience, especially for Korean users and repeated simulation review. Stay with upstream if you need the absolute newest upstream commit and prefer to minimize fork-specific UX changes.
moekyawaung-graduate/MiroFish
Prefer upstream unless you specifically want this fork's repository identity or a minimally modified starting point. This fork is not meaningfully differentiated; it mainly lags one commit behind and adds no visible capabilities.
Choose this fork if you want MiroFish as a more portable, counterfactual-focused simulation platform. Stick with upstream if you want the original Zep-centered path and the most conservative, fully upstream-aligned experience.
fantasyslr/MiroFishmoody
70/100
active
significant_divergence
Prefer this fork if you need a campaign-focused decision system and want the upstream stack adapted for internal evaluation workflows. Prefer upstream if you want the broader prediction/simulation sandbox, faster access to upstream updates, or a more general-purpose demo platform.