Skip to main content
The migration skills in slng-ai/skills move an existing voice agent onto SLNG. Hand one to your coding agent and it discovers your current wiring, swaps speech to text and text to speech for SLNG hosted models, and can point the language model at the Context Router. The rest of your pipeline stays as it is. These skills back the assisted path in the framework guides: in the dashboard you choose your Listen, Think, and Speak models, copy the generated prompt, and the coding agent installs the matching skill and applies the change. For the core skills that call the API directly, see the Agent skill.

Install

The skills CLI needs Node 18 or later. Install a single migration skill by name:
Target one coding agent with -a (claude-code, codex, or gemini-cli), and add -g to install for every project. Re-run the command any time to update.

Set up your key

The migration skills read your key from SLNG_API_KEY:
Create a key at app.slng.ai, or see Create your API key.
The core skills use VOICEAI_API_KEY and the migration skills use SLNG_API_KEY. Set both to the same key if you install skills from each group.

Migration skills

How a migration runs

Once the skill is installed, describe the migration to your coding agent:
The skill guides the agent to:
  1. Discover the entry point, package manager, environment loading, and current speech to text, text to speech, and language model wiring before editing.
  2. Establish a rollback point with a clean git tree.
  3. Validate SLNG_API_KEY without printing it.
  4. Install the framework plugin and replace the speech to text and text to speech constructors with the SLNG equivalents.
  5. Point the language model at the Context Router where you chose to.
  6. Verify one spoken turn that completes speech to text, the language model, and text to speech.

Next steps

LiveKit

Run the assisted or manual migration for a LiveKit agent.

Pipecat

Run the assisted or manual migration for a Pipecat bot.

Custom LLM

Route a custom stack through the Context Router.