Clarifyprompt

ClarifyPrompt MCP

A context-aware MCP prompt compiler that transforms vague prompts into platform-optimized prompts for 60+ AI platforms across 7 categories — grounded in your workspace signals (CLAUDE.md, AGENTS.md, .cursorrules, package.json), resolved intent, and the capabilities of the target model.

Send a raw prompt. ClarifyPrompt gathers the right context, resolves what you're actually trying to do, and returns a version specifically optimized for Midjourney, GPT Image 2, Veo, Runway, Kling, Seedance, Higgsfield, ElevenLabs, Suno, Claude, ChatGPT, Cursor, or any of the 61 supported platforms — with the right syntax, parameters, structure, and grounding.

New in 1.16.0: the platform-currency release. Every built-in pack fact-verified against current vendor docs (2026-10-05) — Midjourney V8.x, Runway Gen-4.5, Veo 3.1, Kling 3.0/Omni, Pika 2.2, Suno v6, ElevenLabs v3, gpt-4o-mini-tts, GPT Image 2 — each entry now carries a verifiedAt date. Sora retired everywhere (OpenAI shut the Videos API 2026-09-24, no replacement); new platforms: Seedance, ElevenLabs Music v2.5, Lyria 3.5. Plus the dialect-currency benchmark: 12 of the newest Ollama Cloud models asked to write platform prompts unprompted — 12/13 emit stale Midjourney flags, 12/12 write fluent prompts for dead Sora, zero hedge. Grounded mode re-runs the same asks with the pack facts injected and the scores flip. See the benchmark and CHANGELOG.md.

New in 1.15.0: Nano Banana (Google Gemini 2.5 Flash Image) is now a built-in image platform — optimize_prompt(platform: "nano-banana") compiles image prompts in its native style (natural-language scene direction, photographic terms, edit-preserving-identity phrasing, in-image text). Plus latest-model compatibility across every provider: claude-sonnet-5, gpt-5/o-series, and Gemini reject temperature and/or max_tokens; the client now sends the right parameters (proactively for known reasoning ids, and learns the rest from a 400). Verified live against Anthropic, OpenAI, Gemini, and Ollama Cloud. See CHANGELOG.md.

New in 1.14.1: Portable-by-default text output — a chat/document/code prompt with no explicit platform now stays platform-neutral instead of quietly defaulting to Claude's idioms (XML tags); name a platform to opt into vendor-specific tuning. Plus the MCP Apps compose panel now shows a for badge and a clean Your prompt → Optimized before/after (with a show changes toggle) instead of an always-on diff. See CHANGELOG.md.

New in 1.14.0: An interactive compose panel via MCP Apps. In hosts that speak the io.modelcontextprotocol/ui extension (Claude Desktop, ChatGPT, Cursor, VS Code, …), compose_prompt renders a live panel: original-vs-optimized view, all six critique scores, the pipeline stages, and Accept / Revise actions — Accept records the outcome into ClarifyPrompt's memory loop, Revise sends your feedback back into the chat. One self-contained ui:// resource; hosts without the extension see zero change. See CHANGELOG.md.

New in 1.13.0: Plain-language rewrites. Optimized prompts now stick to common, everyday words instead of drifting into formal vocabulary ("use", never "utilize") — specificity comes from concrete details, not fancier synonyms. critique_prompt gained a 6th default dimension, `plain_language`, so auto_revise loops correct register drift automatically. Also fixed: an explicit mode (e.g. simple) is no longer silently dropped for small local models under compact system-prompt shaping. See CHANGELOG.md.

The proof this matters — measured, not asserted

Why does this project exist? Because "just ask the model" quietly went stale, and we measured it. Models are frozen at training time; platforms ship new versions monthly (and kill old ones). The result is that asking a modern model for a platform prompt is asking a 2024 encyclopedia for 2026 prices.

On 2026-10-05 we ran a public benchmark: asked 12 of the newest Ollama Cloud models to write prompts for real current platforms, in two conditions, same asks, same models:

• RAW — the model alone, no help. What a user gets in a plain chat.
• GROUNDED — the same asks with the *verified platform facts* injected first — the same pack data ClarifyPrompt injects on every optimize_prompt call. (This is what "grounding" means: the model reads the reference card instead of trusting its frozen memory.)

132 graded answers per condition, verified against vendor docs the same day. Full audit of every raw reply: [`evals/dialect-benchmark/out/final.json`](./evals/dialect-benchmark/out/final.json).

The results, one line:

Same models. Same asks. 35 stale answers → 0.

What that looks like in practice:

• Sora 2, 12 of 12. OpenAI shut the Videos API 2026-09-24 (consumer app April 26). Eleven days later, every model — including GPT-OSS, OpenAI's own open weights — wrote fluent, confident prompts for a product that doesn't exist. No one said so. Not a single hedged answer, not one knowledge-cutoff disclaimer.
• Midjourney, 12 of 13. Everyone reached for --v 6.1. The current default has been V8.2 since June. Two major versions stale, uniformly.
• With the pack facts injected, every model flipped. Same models: all 12 said Sora is discontinued; all 12 pinned --v 8.2; the worst raw performer went 3 current / 5 stale → 8 / 0.

Total cost of both conditions: $0.71. Harness + facts file + grading script: evals/dialect-benchmark/ — reproduce with node evals/dialect-benchmark/run.mjs.

The honest read: models don't *know* current dialects — but with a maintained knowledge layer they execute perfectly. The knowledge is the gap; the pack layer is the product. And that's what optimize_prompt is: grounding, on every call, with packs we refresh against live docs and date with verifiedAt.

How It Works

ClarifyPrompt does two things a plain prompt template can't. Every output below is a real, unedited capture from optimize_prompt run against this repo (see Provenance at the end of this section).

1 — It knows each platform. Same raw prompt, different target, completely different output:

You write:    "a dragon flying over a castle at sunset"

→ Midjourney A colossal, majestic dragon with shimmering scales soaring over a towering medieval stone castle, dramatic sunset sky with vibrant orange and deep purple hues, cinematic fantasy concept art, volumetric lighting, highly detailed --ar 16:9 --v 6.1 --s 250 --q 2

→ DALL-E A majestic dragon with glowing crimson scales soars over a towering medieval stone castle, silhouetted against a vibrant orange and purple sunset sky. Rendered in a high-fantasy digital art style with dramatic, warm lighting and highly detailed textures, wide aspect ratio.

→ Nano Banana A majestic dragon with deep crimson scales and a leathery, bat-like wingspan glides through the warm, golden-hour sky just above a towering medieval castle made of weathered grey stone. ... Frame this as a wide cinematic landscape shot using a 24mm lens at f/8 for deep depth of field, camera positioned at a slightly elevated three-quarter angle... Aspect ratio 16:9.

Midjourney gets --ar/--v/--s/--q flags; DALL-E and Nano Banana get flag-free natural language — and Nano Banana layers in photographic direction (lens, f-stop, camera angle) and explicit mood, its documented style. Same idea, each platform's native dialect.

Notice `--v 6.1` in the Midjourney capture above. That capture is real (2026-06-22) — and the model reached for a version flag that was *already two major versions stale*; Midjourney's current default is V8.2. Model weights are frozen at training time; platform dialects mutate monthly. That gap is exactly what ClarifyPrompt's verified platform packs (verifiedAt-dated, refreshed against current docs) exist to close — see the staleness story.

2 — It knows what you're working on. This is the part a template can't fake. Drop a vague one-liner while editing src/transport.ts *in this very repo*, and the engine grounds it in your real workspace — package.json, git state, the active file — and resolves intent before it shapes the output:

You write:    "add a configurable request timeout to the http transport"

· active file: src/transport.ts · resolved intent: production-code · grounded in: active-file · workspace-meta · git-state · environment · target-model · platform-hints

→ Cursor Implement a configurable request timeout for the HTTP transport in src/transport.ts. Requirements:

1. Add a new environment variable CLARIFYPROMPT_HTTP_TIMEOUT … (default 30000 ms)
2. Apply this timeout to all incoming requests in the streamable-http transport …
3. Preserve existing behavior for stdio and a2a transports … The implementation should be added to the streamable-http section of startTransport(). (excerpted — the full rewrite has 7 numbered requirement groups)

Nothing in that one-line prompt mentioned the CLARIFYPROMPT_HTTP_* naming convention, the startTransport() entry point, or the stdio/a2a transports it must preserve — the engine read those from the active file and package.json and folded them in. That's the difference between *rephrasing a prompt* and *compiling it against context*.

3 — It can run the whole pipeline. clarify → ground/optimize → critique → revise, in one compose_prompt call — see Previously in 1.4.0 — the composable pipeline below.

Provenance. Image outputs captured via glm-5.2:cloud, the grounded code output via qwen3-coder:480b-cloud — both Ollama cloud models served over Ollama's OpenAI-compatible endpoint (LLM_API_URL=http://localhost:11434/v1), run through optimize_prompt against this repo on 2026-06-22 (the Nano Banana capture added 2026-07-03, same glm-5.2:cloud model). ClarifyPrompt is model-agnostic (any OpenAI-compatible API, local or hosted); outputs are model-dependent — yours will differ in wording, not in structure.

What's new in 1.16.0

The platform-currency release. The packs had rotted — Midjourney was still --v 6.1 while the current default is V8.2, the video pack still described "Runway Gen-3" and "Google Veo 2", and the Higgsfield entry still routed to a dead model. Every bu