Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →lingtai-ai/lingtai-preset-skill
Thin dual-axis router for preset questions: which of the 12 TUI-shipped built-in preset templates (provider axis), and which cross-cutting preset lifecycle mechanic (operation axis) — saving, checking availability, activating/refreshing, endpoint/capability facts, or troubleshooting. Read a child only when it is relevant; this does not describe arbitrary saved presets.
Passed 27 of 27 test cases with the skill, and 5 of 27 without it.
and 27 more in the full report
22 fixed5 passed either way0 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
A newer attempt on gemini-3.6-flash on 10/2/2026 did not produce a publishable comparison (not comparable). The snapshot above is the previously measured run.
One of 22 test cases the skill fixed. Same model, same prompt, with and without the skill.
Which direct child path holds the Grok through OpenCode Go route reference?
| Checked | Without skill | With skill |
|---|---|---|
| reference/grok/SKILL.md | Missed | Passed |
The instructions your model reads when the skill loads.
LingTai ships exactly four provider families, one built-in template each.
This router covers exactly the names returned by BuiltinPresets() in
tui/internal/preset/preset.go. It covers TUI-owned template presets only,
not arbitrary saved presets. Route a named-preset question or revision to the
direct child with the matching name under reference/<name>/SKILL.md. That
child owns the family's endpoint/credential facts, exact TUI surfaces, and
named-preset revision instructions.
There are no per-vendor templates. Any other vendor, subscription, gateway,
or account pool is reached through the openai or anthropic family pointed
at that vendor's compatible endpoint (base_url) — never through a
vendor-named provider, which the kernel rejects.
| Name | Provider | Direct child | Route hint |
|---|---|---|---|
| codex | codex | reference/codex/SKILL.md | ChatGPT-account (Codex OAuth) route; the first-run default |
| claude | claude-code | reference/claude/SKILL.md | local Claude Code CLI login ("claude-p") |
| openai | openai | reference/openai/SKILL.md | any OpenAI-compatible endpoint, Chat Completions or Responses; also account pools such as sub2api / subs-pool |
| anthropic | anthropic | reference/anthropic/SKILL.md | any Anthropic Messages-compatible endpoint |
- name: preset-skill-codex
location: reference/codex/SKILL.md
- name: preset-skill-claude
location: reference/claude/SKILL.md
- name: preset-skill-openai
location: reference/openai/SKILL.md
- name: preset-skill-anthropic
location: reference/anthropic/SKILL.md
Do not merge the Codex OAuth catalog, the Claude CLI aliases, and an
arbitrary endpoint's served models. openai and anthropic have no universal
model: the model is free text and must be checked against the configured
endpoint. The exact constructor in preset.go is always the first source to
inspect; picker, capability, and credential surfaces are listed by each
test cases passed, vs 5 / 27 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.