You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Question]: Any plan to follow OpenCode PR #20259 for Kimi system prompt optimization? #2990
changed the title [-][Question]: Any plan to follow OpenCode PR #20259 for Kimi?[/-][+][Question]: Any plan to follow OpenCode PR #20259 for Kimi system prompt optimization?[/+]on Apr 1, 2026
OMO already has Kimi-specific adaptation rather than a direct copy of OpenCode's prompt.
Current source/docs show:
Sisyphus has a Kimi K2.x-native prompt implementation in src/agents/sisyphus/kimi-k2-6.ts.
That prompt is explicitly tuned around K2.x behavior and is separate from the Claude/GPT prompt style.
Atlas also routes Kimi K2.x models to a dedicated Kimi prompt path.
The model matching docs recommend Kimi K2.5/K2.6 as Claude-family alternatives for Sisyphus-style orchestration.
So the answer is effectively yes: OMO follows the same need at the architecture level, but via OMO-specific prompt routing and Kimi-native prompts rather than directly copying the upstream OpenCode change.
Activity
[-][Question]: Any plan to follow OpenCode PR #20259 for Kimi?[/-][+][Question]: Any plan to follow OpenCode PR #20259 for Kimi system prompt optimization?[/+]code-yeongyu commentedon May 15, 2026
[sisyphus-bot]
OMO already has Kimi-specific adaptation rather than a direct copy of OpenCode's prompt.
Current source/docs show:
src/agents/sisyphus/kimi-k2-6.ts.So the answer is effectively yes: OMO follows the same need at the architecture level, but via OMO-specific prompt routing and Kimi-native prompts rather than directly copying the upstream OpenCode change.