Reve AI Image Generation

Generate, edit, and remix images using the Reve AI API. Use when creating images from text prompts, editing existing images with instructions, or combining/remixing multiple reference images. Requires REVE_API_KEY or REVE_AI_API_KEY environment variable.

Security Scan

The skill appears to do what its name says (call a Reve image API via a Bun script), but there are metadata/instruction mismatches and missing declarations that warrant caution before installing.

Purpose & Capability

Name/description (image generation, edit, remix) align with the included script and SKILL.md. However the registry metadata lists no required environment variables or primary credential while the SKILL.md explicitly requires REVE_API_KEY or REVE_AI_API_KEY — an inconsistency that should be resolved.

Instruction Scope

SKILL.md instructions are scoped to generating, editing, and remixing images via the Reve API, and describe only reading input image files and writing output files. They require a Bun runtime and an API key and do not request unrelated system files or services in the prose.

Install Mechanism

No install spec is provided (instruction-only install), which minimizes installer risk. The script requires Bun to run; the Bun requirement is mentioned in SKILL.md but not surfaced in registry requirements. The included TypeScript script (scripts/reve.ts) will be executed by Bun and can perform network I/O — review its contents before running.

Credentials

The skill logically needs one API key (REVE_API_KEY or REVE_AI_API_KEY), which is proportionate. However the registry metadata declares no required env vars or primary credential while SKILL.md requires an API key — this mismatch means automated permission and credential controls may not be correctly applied. The code file should be inspected to confirm it only reads the stated API key and no other secrets or environment variables.

Persistence & Privilege

The skill does not set always:true and there is no explicit disableModelInvocation flag. Default behavior likely allows model-initiated invocation. This is reasonable for a user-invoked image-generation tool, but if you do not want the model to call the skill autonomously, consider requiring disableModelInvocation or user invocation controls.

What to consider before installing

The skill's purpose (call Reve AI to generate/edit images) matches the documentation and bundled script, but the registry metadata omits the required API key declaration and Bun runtime requirement — this mismatch can hide what credentials/tools will be used. Before installing: (1) inspect scripts/reve.ts to confirm it only uses REVE_API_KEY or REVE_AI_API_KEY and does not access other env vars or unexpected endpoints; (2) run it in a sandbox or isolated environment with a test API key and non-sensitive images; (3) ensure Bun is installed from an official source; (4) if you do not want the model to call the skill autonomously, enable disableModelInvocation or require explicit user invocation; (5) prefer skills with a visible homepage/source and clearer metadata. If you cannot inspect the code, treat this skill as higher risk and avoid providing production credentials.

Generate, edit, and remix images using Reve's AI API.

Prerequisites

• Bun runtime
• REVE_API_KEY or REVE_AI_API_KEY environment variable set

Quick Usage

# Generate image from prompt
bun scripts/reve.ts create "A beautiful sunset over mountains" -o sunset.png

# With aspect ratio
bun scripts/reve.ts create "A cat in space" -o cat.png --aspect 16:9

# Edit existing image
bun scripts/reve.ts edit "Add dramatic clouds" -i photo.png -o edited.png

# Remix multiple images
bun scripts/reve.ts remix "Person from <img>0</img> in scene from <img>1</img>" -i person.png -i background.png -o remix.png

Commands

create

Generate a new image from a text prompt.

Options:

• -o, --output FILE — Output file path (default: output.png)
• --aspect RATIO — Aspect ratio: 16:9, 9:16, 3:2, 2:3, 4:3, 3:4, 1:1 (default: 3:2)
• --version VER — Model version (default: latest)

edit

Modify an existing image using text instructions.

Options:

• -i, --input FILE — Input image to edit (required)
• -o, --output FILE — Output file path (default: output.png)
• --version VER — Model version: latest, latest-fast, reve-edit@20250915, reve-edit-fast@20251030

remix

Combine text prompts with reference images. Use <img>N</img> in prompt to reference images by index (0-based).

Options:

• -i, --input FILE — Reference images (can specify multiple, up to 6)
• -o, --output FILE — Output file path (default: output.png)
• --aspect RATIO — Aspect ratio (same options as create)
• --version VER — Model version: latest, latest-fast, reve-remix@20250915, reve-remix-fast@20251030

Constraints

• Max prompt length: 2560 characters
• Max reference images for remix: 6
• Valid aspect ratios: 16:9, 9:16, 3:2, 2:3, 4:3, 3:4, 1:1

Response

The script outputs JSON with generation details:

{
  "output": "path/to/output.png",
  "version": "reve-create@20250915",
  "credits_used": 18,
  "credits_remaining": 982
}

Errors

• 401 — Invalid API key
• 402 — Insufficient credits
• 429 — Rate limited (includes retry_after)
• 422 — Invalid input (prompt too long, bad aspect ratio)