LLM Wiki is a cross-platform desktop application that turns your documents into an organized, interlinked knowledge base — automatically. Instead of traditional RAG (retrieve-and-answer from scratch every time), the LLM incrementally builds and maintains a persistent wiki from your sources。
TypeScript
import { describe, expect, it } from "vitest"
import { resolveIngestReasoning } from "./reasoning-capabilities"
import { getProviderConfig } from "./llm-providers"
import { resolveConfig } from "@/components/settings/preset-resolver"
import type { LlmConfig } from "@/stores/wiki-store"
const openAiConfig: LlmConfig = {
provider: "openai",
apiKey: "k",
model: "gpt-5",
ollamaUrl: "",
customEndpoint: "",
maxContextSize: 128000,
}
describe("ingest reasoning is settable instead of hardcoded off", () => {
it("defaults to off, preserving the behaviour ingest had when it was hardcoded", () => {
expect(resolveIngestReasoning(openAiConfig)).toEqual({ mode: "off" })
})
it("uses the configured mode when the user picks one", () => {
expect(resolveIngestReasoning({ ...openAiConfig, ingestReasoning: { mode: "low" } }))
.toEqual({ mode: "low" })
})
it("keeps chat reasoning independent of ingest reasoning", () => {
const config: LlmConfig = {
...openAiConfig,
reasoning: { mode: "high" },
ingestReasoning: { mode: "off" },
}
expect(resolveIngestReasoning(config)).toEqual({ mode: "off" })
expect(config.reasoning).toEqual({ mode: "high" })
})
it("leaves provider normalization to the wire builder", () => {
const autoOnly: LlmConfig = {
...openAiConfig,
provider: "custom",
customEndpoint: "https://example.invalid/v1",
}
expect(resolveIngestReasoning(autoOnly)).toEqual({ mode: "off" })
})
it("reaches the wire: the selected effort ends up in the request body", () => {
const withLow = { ...openAiConfig, ingestReasoning: { mode: "low" as const } }
const body = getProviderConfig(withLow).buildBody(
[{ role: "user", content: "hi" }],
{ reasoning: resolveIngestReasoning(withLow) },
) as Record<string, unknown>
expect(body.reasoning_effort).toBe("low")
})
it("survives the preset resolver, which drops fields it does not carry", () => {
const resolved = resolveConfig(
{ id: "openai", label: "OpenAI", provider: "openai" } as never,
{ ingestReasoning: { mode: "low" } },
openAiConfig,
)
expect(resolved.ingestReasoning).toEqual({ mode: "low" })
expect(resolveIngestReasoning(resolved)).toEqual({ mode: "low" })
})
})