# PennyroyalTea/gibberlink

Two conversational AI agents switching from English to sound-level protocol after confirming they are both AI agents

Repository: https://github.com/PennyroyalTea/gibberlink
Canonical: https://ross.abutalabs.com/products/gibberlink
Homepage: https://www.gbrl.ai/
Language: TypeScript
License: MIT
License Family: permissive
Last push: 2025-07-28T06:49:52+00:00

## Health v2 (maintenance only)
Score: 35/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 34, release rhythm 35, longevity 39
- inputs: {"age_days": 556, "days_push": 401, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4887, forks 395 (observed 2026-08-28T04:09:01.914679+00:00)

## What it is
GibberLink is a viral demo application in which two conversational AI agents (built on ElevenLabs) detect that they are both AI and switch from spoken English to the ggwave data-over-sound protocol for more efficient communication. The repository provides the API layer that lets agents emit and decode the sound-level protocol, with a browser-based demo at gbrl.ai.

## Use cases
- let ai agents talk to each other over audio instead of english speech
- demo agents switching from voice to a data-over-sound protocol
- run agent-to-agent conversations in the browser using two devices
- make conversational chatbots detect each other and switch to ggwave
- encode and decode messages between agents over sound
- experiment with protocol negotiation between voice ai agents

## When to choose
- you want a striking, working demo of agent-to-agent acoustic communication
- you are exploring ggwave-based data-over-sound on top of conversational ai agents
- you need a browser-based showcase that judges or audiences can experience without setup
- you want to build derived work on top of a proven viral agent-communication demo

## When to avoid
- you need production-grade, reliable data transfer in noisy real-world environments
- you want a full multi-agent orchestration framework rather than a demo with an api layer
- you require long-distance or high-bandwidth machine-to-machine communication
- you need agents to communicate over networks rather than through a speaker and microphone

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, chatbot, audio-processing, speech-recognition
- domain: artificial-intelligence
- platform: browser
- tags: ggwave, data-over-sound, agent-to-agent, voice-agents, elevenlabs, hackathon-winner, demo, acoustic-communication, ai-agents, audio, natural-language-processing, nodejs, web-server

## Member repositories
- PennyroyalTea/gibberlink (main) score 35

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:01.914679+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:18:13.302619+00:00, confidence not recorded.
  - readme: https://github.com/PennyroyalTea/gibberlink (fetched 2026-08-28T04:09:01.914679+00:00, sha c28730cc00af)
  - homepage: https://www.gbrl.ai/ (fetched 2026-08-29T09:00:52.446027+00:00, sha 766bae365ba0)
- Data as of 2026-08-30T08:39:29.467469+00:00.