@gentbajko/slopify 0.7.0 → 0.8.1

package/dist/adapters/llm/codex.js

@@ -25,6 +25,9 @@ export function codexArgs(req) {
         "--skip-git-repo-check",
         "-c",
         `web_search="${req.webSearch === true ? "live" : "disabled"}"`,
+        ...(req.thinkingConfig?.effort
+            ? ["-c", `model_reasoning_effort="${req.thinkingConfig.effort}"`]
+            : []),
         ...(req.model === "" ? [] : ["-m", req.model]),
         // The prompt is one argv element after `--`, so a leading dash is text, not a flag.
         "--",

package/dist/adapters/llm/gemini-workspace.js

@@ -4,7 +4,7 @@ import { join } from "node:path";
 // Gemini 0.16 supports stream-json and tools.core, but headless default mode still
 // exposes file-reading tools. Isolate each writing call while retaining ~/.gemini
 // authentication. Nothing is written into the user's settings or project.
-export function geminiWorkspace(webSearch) {
+export function geminiWorkspace(webSearch, request) {
     const directory = mkdtempSync(join(tmpdir(), "slopify-gemini-"));
     try {
         const settings = join(directory, "settings.json");
@@ -12,6 +12,29 @@ export function geminiWorkspace(webSearch) {
         const trust = join(directory, "trusted-folders.json");
         const mcpAllowlist = `slopify-disabled-${directory.split(/[\\/]/).at(-1)}`;
         writeFileSync(settings, JSON.stringify({
+            ...(request?.thinkingConfig
+                ? {
+                    modelConfigs: {
+                        customOverrides: [
+                            {
+                                match: { model: request.model },
+                                modelConfig: {
+                                    generateContentConfig: {
+                                        thinkingConfig: {
+                                            ...(request.thinkingConfig.level === undefined
+                                                ? {}
+                                                : { thinkingLevel: request.thinkingConfig.level }),
+                                            ...(request.thinkingConfig.budget === undefined
+                                                ? {}
+                                                : { thinkingBudget: request.thinkingConfig.budget }),
+                                        },
+                                    },
+                                },
+                            },
+                        ],
+                    },
+                }
+                : {}),
             tools: {
                 core: webSearch ? ["google_web_search"] : [],
                 allowed: webSearch ? ["google_web_search"] : [],

package/dist/adapters/llm/gemini.js

@@ -43,7 +43,7 @@ export function geminiLlm(deps) {
     const binary = deps.binary ?? geminiBinary;
     async function* complete(req) {
         req.signal.throwIfAborted();
-        const workspace = geminiWorkspace(req.webSearch === true);
+        const workspace = geminiWorkspace(req.webSearch === true, req);
         let run;
         try {
             run = deps.run(binary, geminiArgs(req, workspace.mcpAllowlist), req.signal, workspace.options);

package/dist/adapters/llm/openrouter.js

@@ -47,6 +47,7 @@ export function openRouterLlm(deps) {
                     role: message.role,
                     content: message.content,
                 })),
+                ...(req.thinkingConfig?.effort ? { reasoning: { effort: req.thinkingConfig.effort } } : {}),
                 stream: true,
                 // The usage-accounting flag: without it the final chunk carries no token counts
                 // and the Usage page would have nothing to count.

package/dist/assets/models.yaml

@@ -0,0 +1,470 @@
+# Curated adapter-compatible models. Edit this file to change the published defaults.
+# A private copy in the Slopify data directory reloads automatically.
+# Omitted pricing means unknown, never free. Provider concurrency is capped at five.
+schemaVersion: 1
+updatedAt: 2026-09-10
+providers:
+  openrouter:
+    maxConcurrent: 3
+  claude-code:
+    maxConcurrent: 3
+  codex:
+    maxConcurrent: 3
+  gemini:
+    maxConcurrent: 3
+  elevenlabs:
+    maxConcurrent: 3
+  openai-tts:
+    maxConcurrent: 3
+  cartesia:
+    maxConcurrent: 3
+  inworld:
+    maxConcurrent: 5
+  fal:
+    maxConcurrent: 3
+  replicate:
+    maxConcurrent: 3
+  openai-image:
+    maxConcurrent: 3
+  google-image:
+    maxConcurrent: 3
+llm:
+  - provider: claude-code
+    id: opus
+    name: Claude Opus (latest)
+    source: https://code.claude.com/docs/en/model-config
+    keywords:
+      - llm
+    pricing: &a1
+      note: Uses your CLI login and account plan. No per-run price is available; paid
+        overage may apply.
+    llm:
+      webSearch: true
+  - provider: claude-code
+    id: sonnet
+    name: Claude Sonnet (latest)
+    source: https://code.claude.com/docs/en/model-config
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+  - provider: claude-code
+    id: haiku
+    name: Claude Haiku (latest)
+    source: https://code.claude.com/docs/en/model-config
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+  - provider: codex
+    id: gpt-6-astra
+    name: GPT-6-Astra
+    source: https://developers.openai.com/codex/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: &a2
+        low:
+          effort: low
+        medium:
+          effort: medium
+        high:
+          effort: high
+  - provider: codex
+    id: gpt-5.6-sol
+    name: GPT-5.6-Sol
+    source: https://developers.openai.com/codex/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: *a2
+  - provider: codex
+    id: gpt-5.6-terra
+    name: GPT-5.6-Terra
+    source: https://developers.openai.com/codex/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: *a2
+  - provider: codex
+    id: gpt-5.6-luna
+    name: GPT-5.6-Luna
+    source: https://developers.openai.com/codex/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: *a2
+  - provider: codex
+    id: gpt-5.5
+    name: GPT-5.5
+    source: https://developers.openai.com/codex/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: *a2
+  - provider: codex
+    id: gpt-5.3-codex-spark
+    name: GPT-5.3-Codex-Spark
+    source: https://developers.openai.com/codex/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: *a2
+  - provider: gemini
+    id: gemini-3.8-flash
+    name: Gemini 3.8 Flash
+    source: https://ai.google.dev/gemini-api/docs/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: &a3
+        low:
+          level: low
+        medium:
+          level: medium
+        high:
+          level: high
+  - provider: gemini
+    id: gemini-3.1-pro-preview
+    name: Gemini 3.1 Pro (preview)
+    source: https://ai.google.dev/gemini-api/docs/models
+    keywords:
+      - llm
+    pricing: *a1
+    llm:
+      webSearch: true
+      thinking: *a3
+  - provider: openrouter
+    id: google/gemini-3.8-flash
+    name: "Google: Gemini 3.8 Flash"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 0.75
+      outputPerMillionTokens: 3.75
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1048576
+      thinking: *a2
+  - provider: openrouter
+    id: google/gemini-3.1-pro-preview
+    name: "Google: Gemini 3.1 Pro Preview"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 2
+      outputPerMillionTokens: 12
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1048576
+      thinking: *a2
+  - provider: openrouter
+    id: openai/gpt-5.6-sol
+    name: "OpenAI: GPT-5.6 Sol"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 2
+      outputPerMillionTokens: 10
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1050000
+      thinking: *a2
+  - provider: openrouter
+    id: openai/gpt-5.6-terra
+    name: "OpenAI: GPT-5.6 Terra"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 2
+      outputPerMillionTokens: 12
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1050000
+      thinking: *a2
+  - provider: openrouter
+    id: openai/gpt-5.6-luna
+    name: "OpenAI: GPT-5.6 Luna"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 0.19999999999999998
+      outputPerMillionTokens: 1.2
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1050000
+      thinking: *a2
+  - provider: openrouter
+    id: anthropic/claude-sonnet-4.6
+    name: "Anthropic: Claude Sonnet 4.6"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 3
+      outputPerMillionTokens: 15
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1000000
+      thinking: *a2
+  - provider: openrouter
+    id: anthropic/claude-opus-4.7
+    name: "Anthropic: Claude Opus 4.7"
+    source: https://openrouter.ai/api/v1/models
+    keywords:
+      - llm
+    pricing:
+      inputPerMillionTokens: 5
+      outputPerMillionTokens: 25
+      note: Public OpenRouter rates verified 2026-09-10; routing, long context,
+        reasoning and web search may add charges.
+    llm:
+      webSearch: true
+      contextTokens: 1000000
+      thinking: *a2
+image:
+  - provider: openai-image
+    id: gpt-image-2.5-sunburst
+    name: GPT Image 2.5 Sunburst
+    source: https://developers.openai.com/api/docs/guides/image-generation
+    keywords:
+      - image
+    pricing:
+      note: Automatic quality and generated image tokens determine the charge; a fixed
+        image price is unavailable.
+    image: &a4
+      aspectRatios: &a5
+        - 16:9
+        - 9:16
+  - provider: openai-image
+    id: gpt-image-2.5-flare
+    name: GPT Image 2.5 Flare
+    source: https://developers.openai.com/api/docs/guides/image-generation
+    keywords:
+      - image
+    pricing:
+      note: Automatic quality and generated image tokens determine the charge; a fixed
+        image price is unavailable.
+    image: *a4
+  - provider: openai-image
+    id: gpt-image-2
+    name: GPT Image 2
+    source: https://developers.openai.com/api/docs/guides/image-generation
+    keywords:
+      - image
+    pricing:
+      note: Automatic quality and generated image tokens determine the charge; a fixed
+        image price is unavailable.
+    image: *a4
+  - provider: google-image
+    id: gemini-3.1-flash-image
+    name: Nano Banana 2
+    source: https://ai.google.dev/gemini-api/docs/pricing
+    keywords:
+      - image
+    pricing:
+      perImage: 0.101
+      note: 2K image output price; prompt, thinking and other input tokens are
+        additional.
+    image:
+      aspectRatios: *a5
+      resolution: 2K
+  - provider: google-image
+    id: gemini-3-pro-image
+    name: Nano Banana Pro
+    source: https://ai.google.dev/gemini-api/docs/models
+    keywords:
+      - image
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    image:
+      aspectRatios: *a5
+      resolution: 2K
+  - provider: fal
+    id: fal-ai/flux-2
+    name: FLUX.2
+    source: https://fal.ai/models/fal-ai/flux-2
+    keywords:
+      - image
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    image: *a4
+  - provider: fal
+    id: fal-ai/nano-banana-2
+    name: Nano Banana 2
+    source: https://fal.ai/models/fal-ai/nano-banana-2
+    keywords:
+      - image
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    image: *a4
+  - provider: replicate
+    id: black-forest-labs/flux-1.1-pro
+    name: FLUX 1.1 Pro
+    source: https://replicate.com/black-forest-labs/flux-1.1-pro
+    keywords:
+      - image
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    image: *a4
+  - provider: replicate
+    id: black-forest-labs/flux-dev
+    name: FLUX.1 Dev
+    source: https://replicate.com/black-forest-labs/flux-dev
+    keywords:
+      - image
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    image: *a4
+  - provider: replicate
+    id: black-forest-labs/flux-schnell
+    name: FLUX.1 Schnell
+    source: https://replicate.com/black-forest-labs/flux-schnell
+    keywords:
+      - image
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    image: *a4
+tts:
+  - provider: elevenlabs
+    id: eleven_v3
+    name: Eleven v3
+    source: https://elevenlabs.io/docs/overview/models
+    keywords:
+      - tts
+    pricing:
+      perMillionCharacters: 100
+      note: Advertised starting API rate; your plan, voice and account credits affect
+        actual cost.
+    tts:
+      maxCharacters: 5000
+      streaming: true
+  - provider: elevenlabs
+    id: eleven_multilingual_v2
+    name: Eleven Multilingual v2
+    source: https://elevenlabs.io/docs/overview/models
+    keywords:
+      - tts
+    pricing:
+      perMillionCharacters: 100
+      note: Advertised starting API rate; your plan, voice and account credits affect
+        actual cost.
+    tts:
+      maxCharacters: 10000
+      streaming: true
+  - provider: elevenlabs
+    id: eleven_flash_v2_5
+    name: Eleven Flash v2.5
+    source: https://elevenlabs.io/docs/overview/models
+    keywords:
+      - tts
+    pricing:
+      perMillionCharacters: 50
+      note: Advertised starting API rate; your plan, voice and account credits affect
+        actual cost.
+    tts:
+      maxCharacters: 40000
+      streaming: true
+  - provider: inworld
+    id: inworld-tts-2
+    name: Inworld TTS-2
+    source: https://inworld.ai/pricing
+    keywords:
+      - tts
+    pricing:
+      perMillionCharacters: 25
+      note: On-Demand rate before discounts or credits. Requests are split at 10,000
+        characters for account compatibility.
+    tts:
+      maxCharacters: 10000
+      asyncMaxCharacters: 100000
+      streaming: true
+  - provider: inworld
+    id: inworld-tts-2-flash
+    name: Inworld TTS-2 Flash
+    source: https://inworld.ai/pricing
+    keywords:
+      - tts
+    pricing:
+      perMillionCharacters: 15
+      note: On-Demand rate before plan discounts or credits.
+    tts:
+      maxCharacters: 4000
+      streaming: true
+  - provider: cartesia
+    id: sonic-3.6
+    name: Sonic 3.6
+    source: https://docs.cartesia.ai/build-with-cartesia/tts-models/latest
+    keywords:
+      - tts
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    tts:
+      maxCharacters: 4000
+      streaming: true
+  - provider: openai-tts
+    id: gpt-4o-mini-tts
+    name: GPT-4o mini TTS
+    source: https://developers.openai.com/api/docs/guides/text-to-speech
+    keywords:
+      - tts
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    tts:
+      maxCharacters: 4096
+      streaming: true
+  - provider: openai-tts
+    id: tts-1
+    name: TTS-1
+    source: https://developers.openai.com/api/docs/guides/text-to-speech
+    keywords:
+      - tts
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    tts:
+      maxCharacters: 4096
+      streaming: true
+  - provider: openai-tts
+    id: tts-1-hd
+    name: TTS-1 HD
+    source: https://developers.openai.com/api/docs/guides/text-to-speech
+    keywords:
+      - tts
+    pricing:
+      note: Price depends on your account or generated output; estimate unavailable.
+    tts:
+      maxCharacters: 4096
+      streaming: true

package/dist/catalog/registry.js

@@ -0,0 +1,147 @@
+import { providerError } from "../kernel/ports/model.js";
+import { splitText } from "../kernel/ports/text.js";
+export function curateRegistry(registry, catalogue) {
+    const requireModel = (provider, family, id) => {
+        const models = catalogue.models(provider, family);
+        const model = id === undefined ? models[0] : models.find((m) => m.id === id);
+        if (!model)
+            throw providerError({
+                kind: "unsupported",
+                message: `${provider}: this model is not enabled in the current catalogue. Choose a supported model in Run settings.`,
+            });
+        return model;
+    };
+    return {
+        list: registry.list,
+        llm: (id) => {
+            const port = registry.llm(id);
+            return {
+                ...port,
+                models: async () => catalogue.models(id, "llm").map((m) => ({
+                    ...m,
+                    ...("llm" in m && m.llm.thinking
+                        ? {
+                            thinkingModes: Object.keys(m.llm.thinking).filter((mode) => ["off", "low", "medium", "high", "xhigh"].includes(mode)),
+                        }
+                        : {}),
+                })),
+                complete: async function* (request) {
+                    const model = requireModel(id, "llm", request.model);
+                    if (!("llm" in model))
+                        throw new Error("Invalid LLM catalogue entry");
+                    if (request.webSearch && !model.llm.webSearch)
+                        throw providerError({
+                            kind: "unsupported",
+                            message: "This model does not support research web search.",
+                        });
+                    const thinkingConfig = request.thinking === undefined ? undefined : model.llm.thinking?.[request.thinking];
+                    if (request.thinking !== undefined && !thinkingConfig)
+                        throw providerError({
+                            kind: "unsupported",
+                            message: "This model does not support the selected thinking setting. Choose an available setting.",
+                        });
+                    yield* port.complete({ ...request, ...(thinkingConfig ? { thinkingConfig } : {}) });
+                },
+            };
+        },
+        image: (id) => {
+            const port = registry.image(id);
+            return {
+                ...port,
+                models: async () => catalogue.models(id, "image"),
+                generate: async (request) => {
+                    const model = requireModel(id, "image", request.model);
+                    if (!("image" in model) || !model.image.aspectRatios.includes(request.aspect))
+                        throw providerError({
+                            kind: "unsupported",
+                            message: "This model does not support the selected video shape.",
+                        });
+                    return port.generate(request);
+                },
+            };
+        },
+        tts: (id) => {
+            const port = registry.tts(id);
+            return {
+                ...port,
+                models: async () => catalogue.models(id, "tts"),
+                synthesize: async (request) => {
+                    const model = requireModel(id, "tts", request.model);
+                    if (!("tts" in model))
+                        throw new Error("Invalid TTS catalogue entry");
+                    const parts = splitText(request.text, model.tts.maxCharacters);
+                    if (parts.length <= 1)
+                        return port.synthesize({ ...request, model: model.id });
+                    const abort = new AbortController();
+                    const signal = AbortSignal.any([request.signal, abort.signal]);
+                    const iterator = audioParts(parts, { ...request, model: model.id, signal }, port.synthesize);
+                    return {
+                        container: "mp3",
+                        audio: new ReadableStream({
+                            async pull(controller) {
+                                try {
+                                    const next = await iterator.next();
+                                    if (next.done)
+                                        controller.close();
+                                    else
+                                        controller.enqueue(next.value);
+                                }
+                                catch (error) {
+                                    controller.error(error);
+                                }
+                            },
+                            async cancel() {
+                                abort.abort();
+                                await iterator.return();
+                            },
+                        }, { highWaterMark: 0 }),
+                    };
+                },
+            };
+        },
+    };
+}
+async function* audioParts(parts, request, synthesize) {
+    let tokens = {};
+    const saved = request.continuation?.read();
+    if (saved) {
+        try {
+            const parsed = JSON.parse(saved);
+            if (parsed && typeof parsed === "object" && !Array.isArray(parsed))
+                tokens = Object.fromEntries(Object.entries(parsed).filter((entry) => typeof entry[1] === "string"));
+        }
+        catch {
+            /* A legacy single-job token cannot identify several split jobs. */
+        }
+    }
+    for (const [index, text] of parts.entries()) {
+        request.signal.throwIfAborted();
+        // Body chunks are persisted individually by narration. Long intro/outro text
+        // is bounded here as well; each split has its own remote continuation.
+        const response = await synthesize({
+            ...request,
+            text,
+            continuation: {
+                read: () => tokens[String(index)],
+                write: (value) => {
+                    tokens[String(index)] = value;
+                    request.continuation?.write(JSON.stringify(tokens));
+                },
+            },
+        });
+        const reader = response.audio.getReader();
+        try {
+            for (;;) {
+                const next = await reader.read();
+                request.signal.throwIfAborted();
+                if (next.done)
+                    break;
+                yield next.value;
+            }
+        }
+        finally {
+            await reader.cancel().catch(() => { });
+            reader.releaseLock();
+        }
+    }
+}