This release adds 1 notable feature for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Summary
AI summaryUpdates packs/platforms/image.yaml + video.yaml, Numbers, and Compatibility across a mixed release.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Low |
Adds Higgsfield as a target platform for image and video categories. Adds Higgsfield as a target platform for image and video categories. Source: llm_adapter@2026-05-26 Confidence: high |
— |
| Feature | Low |
Introduces syntax hints for Higgsfield model identifiers and prompt conventions in YAML files. Introduces syntax hints for Higgsfield model identifiers and prompt conventions in YAML files. Source: granite4.1:30b@2026-05-26-audit Confidence: low |
— |
Full changelog
Summary
Patch release. Adds Higgsfield as a target platform in both image and video categories. Pure YAML platform-pack additions; no code changes, no MCP tool surface changes.
What Higgsfield brings
Higgsfield is a multi-model creative platform that exposes its own hosted MCP server at https://mcp.higgsfield.ai/mcp (9 tools — generate_image, generate_video, show_characters, virality_predictor, etc. Auth via Higgsfield account; no API key). Inside one connection it routes to 30+ underlying models:
- Image: Soul 2.0, Soul Cinema, Soul Cast (character-consistent), Flux 2, Seedream 5, Nano Banana Pro, GPT Image 2
- Video: Cinema Studio (Higgsfield-original), Sora 2, Veo 3.1, Kling 3.0, WAN 2.6, Seedance 2.0
- Workflows: Soul ID character training, Lipsync Studio, UGC Factory, Marketing Studio (virality predictor)
What 1.6.1 adds
The 1.6.1 ClarifyPrompt platform entries surface Higgsfield's model identifiers + prompt-style conventions (long-form natural-language prose; composition + lighting + textures + mood; up to 4K images / 15 s video / Soul ID for character consistency) as syntax hints to the curator.
# packs/platforms/image.yaml + video.yaml
- id: higgsfield
label: Higgsfield (multi-model platform)
description: ...
syntaxHints:
- "models: soul-2.0 | flux-2 | seedream-5 | nano-banana-pro | ..."
- long-form natural-language prose (composition + lighting + textures + mood)
- ...
Recommended composition pattern
Install BOTH clarifyprompt-mcp AND Higgsfield's MCP in your client (Claude Desktop / Cursor / AI Butler / Claude Code). Use ClarifyPrompt to compile, then pass to Higgsfield to generate:
optimize_prompt(platform: 'higgsfield', prompt: '...')
↓ (compiled with Higgsfield's prose conventions)
higgsfield.generate_image(prompt: '...', model: 'soul-2.0')
MCPs compose at the client; ClarifyPrompt stays at the compile layer.
Numbers
- Platforms: 58+ → 60+ (Image 10 → 11, Video 11 → 12)
- Fixtures: 29 → 30 (added
30-higgsfield-platform-loaded.yaml) - Local eval baseline: 26 passed / 1 failed (known
analyzer-creative-media) / 3 skipped / 97% avg on qwen-coder-7b - Tarball: ~152 kB packed, +2 KB for the YAML entries
Compatibility
- No code changes. No MCP tool surface changes (still 23 tools).
- No env-var changes.
- Fully back-compat with 1.6.0.
Full diff: https://github.com/LumabyteCo/clarifyprompt-mcp/compare/v1.6.0...v1.6.1
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About LumabyteCo/clarifyprompt-mcp
MCP server for AI prompt optimization — transforms vague prompts into platform-optimized prompts for 58+ AI platforms across 7 categories (image, video, voice, music, code, chat, document).
Related context
Beta — feedback welcome: [email protected]