A research project.
To track the responses of inquiries into system prompts from the main ai llm models over time.
- Runs daily GitHub Actions workflow
- Fetches system prompts from various AI models and see what we get
- Archives prompts with timestamps in a structured format
- Maintains history of changes over time in the repo publically
- No API keys or private credentials stored in the repository
- All data sourced from official public channels
Want more models? More detail?
- UI that pulls in all the log data to visualise, diff, etc.
- randomisation of prompts to a/b test
- follow up questions
- ...etc
idea: https://x.com/NickADobos/status/1893872882658193729
inspo: https://github.com/statsig-io/statuspage
leave an issue here, or dm me
Get your api gets for providers here:
- https://console.anthropic.com/settings/keys
- https://platform.openai.com/api-keys
- https://console.x.ai/
- https://aistudio.google.com/app/apikey
