Six foundational questions define GitHub Minimum Intelligence. Each examines the project from a different angle—together they form a complete picture of what repository-native intelligence means and why it matters.
| Question | Summary |
|---|---|
| What? | Defines GitHub Minimum Intelligence as a repository-native AI collaboration framework—not a hosted platform, but a local intelligence layer built from issues, workflows, markdown, and commits. |
| Who? | Explores identity and accountability: who speaks, who executes, who remembers, and who governs when the repository itself becomes the mind. |
| When? | Examines the temporal architecture: how Git commits replace ephemeral sessions, making memory durable, trust auditable, and collaboration resilient across time. |
| Where? | Maps where intelligence physically lives—runtime in GitHub Actions, memory in versioned state files, identity in checked-in markdown, and authorization in existing repository permissions. |
| How? | Walks through the implementation—issues as conversational input, Actions as execution runtime, an LLM as reasoning substrate, and Git commits as durable memory. |
| How Much? | Asks how much intelligence a repository can hold, and argues the ceiling is social stewardship, not token count—memory scales with Git history, not context windows. |