← Back to Skills Marketplace
517
Downloads
1
Stars
9
Active Installs
5
Versions
Install in OpenClaw
/install pollinations-ai
Description
Generate images, music, and videos from text prompts using Pollinations AI with models like flux, zimage, and suno-4 via API key.
Usage Guidance
This skill appears to do what it says: call Pollinations API to generate images, audio, and video. Before installing: (1) only provide a Pollinations API key — avoid giving unrelated secrets; (2) note the optional Telegram feature uses an external 'openclaw' CLI that is not declared in the metadata, so Telegram sending will fail unless you have an appropriate 'openclaw' command installed and trust it; (3) verify ALLOWED_OUTPUT_DIRS and OPENCLAW_WORKSPACE values so generated files are written only where you expect (SKILL.md mentions '/tmp' as a default but the code primarily uses the workspace outputs directory); (4) review network policy if you need to restrict outbound calls to gen.pollinations.ai. If any of the undocumented behaviors (use of 'openclaw' or differences in ALLOWED_OUTPUT_DIRS defaults) are unacceptable, ask the author to update metadata and docs before use.
Capability Analysis
Type: OpenClaw Skill
Name: pollinations-ai
Version: 1.0.4
The pollinations-ai skill bundle is a legitimate tool for generating AI media via the Pollinations AI API. It includes proactive security measures like path traversal sanitization in `pollinations_image_gen.py`, `pollinations_audio_gen.py`, and `pollinations_video_gen.py`, and avoids hardcoding credentials by using environment variables. The use of `subprocess.run` to interact with the OpenClaw CLI for Telegram messaging is implemented safely without shell execution, and the instructions in `SKILL.md` are consistent with the stated functionality.
Capability Assessment
Purpose & Capability
Name/description match the code: three CLI scripts call Pollinations endpoints to generate images, audio, and video. The only capability that extends beyond pure generation is an optional Telegram send implemented via an 'openclaw message send' subprocess call — the skill does not declare 'openclaw' as a required binary in metadata, so that dependency is undocumented.
Instruction Scope
SKILL.md instructions align with the code: require POLLINATIONS_API_KEY, explain optional TELEGRAM_CHAT_ID, and show how to run the scripts. The scripts only read the declared env vars (POLLINATIONS_API_KEY, TELEGRAM_CHAT_ID, ALLOWED_OUTPUT_DIRS, OPENCLAW_WORKSPACE) and write output files to allowed directories according to sanitizers. No unexpected file reads, secret harvesting, or external endpoints beyond gen.pollinations.ai are present.
Install Mechanism
No install spec is provided (instruction-only install). The included requirements.txt and meta.json list reasonable Python dependencies (requests, Pillow). There are no downloads from arbitrary URLs or archive extraction.
Credentials
Only POLLINATIONS_API_KEY is required (with TELEGRAM_CHAT_ID, ALLOWED_OUTPUT_DIRS, OPENCLAW_WORKSPACE optional). This is proportionate to the stated functionality. The scripts avoid hardcoding secrets and mask the API key in logged URLs.
Persistence & Privilege
The skill does not request permanent 'always' inclusion and does not modify other skills or system-wide settings. It creates output files under a workspace but does not persist credentials or change agent configuration.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install pollinations-ai - After installation, invoke the skill by name or use
/pollinations-ai - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.4
- Updated default video model from "suno" to "suno-4" in the video generation section.
- No functionality or parameter changes; documentation update only.
v1.0.3
- Fixed default video model to suno-4 in CLI documentation for pollinations_video_gen.py.
- No functional changes to scripts; documentation alignment only.
- Updated SKILL.md to reflect accurate video generation defaults.
v1.0.2
Version 1.0.2 – Documentation & Security Updates
- Major update to SKILL.md with expanded setup instructions, improved usage guides, and full parameter documentation.
- Added detailed security configuration and environment variable documentation, clarifying required and optional settings.
- Outlined built-in security/mechanisms (allowed output directories, Telegram opt-in).
- Enhanced references, troubleshooting, and supported model listings for clarity.
- Added _meta.json for improved metadata management.
v1.0.1
- API Key is now required for all models.
- Updated documentation for simplified setup instructions.
- Clarified recommended and available models list.
- Improved CLI usage and parameters section.
- Removed the detailed models reference file.
v1.0.0
Initial release of pollinations-ai skill.
- Generate images, music, and video from text prompts using the Pollinations AI API.
- Supports multiple image, audio, and video models (e.g., flux, zimage, nanobanana, suno-4).
- Includes scripts for dedicated image, audio, and video generation.
- Allows customization of model, output size, duration, aspect ratio, and other parameters.
- Offers step-by-step setup instructions for API key and dependencies.
- Provides troubleshooting guide and integration example for use in OpenClaw.
Metadata
Frequently Asked Questions
What is Pollinations AI?
Generate images, music, and videos from text prompts using Pollinations AI with models like flux, zimage, and suno-4 via API key. It is an AI Agent Skill for Claude Code / OpenClaw, with 517 downloads so far.
How do I install Pollinations AI?
Run "/install pollinations-ai" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Pollinations AI free?
Yes, Pollinations AI is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Pollinations AI support?
Pollinations AI is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Pollinations AI?
It is built and maintained by kanfred (@kanfred); the current version is v1.0.4.
More Skills