Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →lingtai-ai/preset-skill-minimax
Official-source-led manual for the TUI `minimax` template.
Passed 17 of 21 test cases with the skill, and 8 of 21 without it.
and 9 more in the full report
12 fixed8 passed either way3 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 12 test cases the skill fixed. Same model, same prompt, with and without the skill.
Which model identifiers constitute the complete native CN dropdown options in providerModels['minimax'] within preset_editor.go?
| Checked | Without skill | With skill |
|---|---|---|
| Configures the native CN picker with MiniMax-M2.7, MiniMax-M2.7-highspeed, MiniMax-M2.5, and MiniMax-M2.5-highspeed | Missed | Passed |
The instructions your model reads when the skill loads.
Use this child for the named built-in minimax preset. Its current constructor is minimaxPreset in tui/internal/preset/preset.go:1248: it ships MiniMax-M2.7, MINIMAX_API_KEY, the CN Anthropic-compatible default, and text/tool capabilities. The regional rows are ProviderRegionURLs["minimax"]: CN, INTL, and OpenCode Go.
Read the official MiniMax Models catalog and its API overview. For served IDs, call the selected official Anthropic route's authenticated GET /anthropic/v1/models; use the catalog/API docs for modality and plan facts. OpenCode Go is a gateway, so separately query its authenticated GET https://opencode.ai/zen/go/v1/models and do not copy a MiniMax catalog into the Go row. Preserve exact case and distinguish the Go credential from the regional slots. Check the official Token Plan MCP guide only for its separate manual image tool; it does not make the TUI vision capability change. The official MiniMax-Coding-Plan-MCP repository ships the minimax-coding-plan-mcp package and understand_image tool for a Token Plan seat or purchased Credits; it is manual-only and not auto-wired here.
Start at minimaxPreset in tui/internal/preset/preset.go. For a model change, revise providerModels["minimax"] and the matching modelHasVision entries in tui/internal/tui/preset_editor.go; for endpoint or credential changes, revise ProviderRegionURLs and ProviderDefaultEnv in preset.go. Confirm the fixed capability rendering only if the constructor's vision wiring changes. Follow the curation and retired-model rules in tui/internal/tui/SKILL.md and tui/CONTRACT.md.
The native CN picker is MiniMax-M2.7, MiniMax-M2.7-highspeed, MiniMax-M2.5, and MiniMax-M2.5-highspeed, all explicitly text-only for the reviewed Anthropic-compatible route. INTL remains free text because native parity is not proven. OpenCode Go keeps its exact pre-PR picker — MiniMax-M3, MiniMax-M2.7, and MiniMax-M2.7-highspeed — and its prior behavior; native modality metadata is not copied into that gateway route. Retired IDs remain valid only when already pinned in a saved preset.
test cases passed, vs 8 / 21 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.