llms.txt

# Builder Weekly Tutorials
# Agent-building patterns where AI agents measurably beat humans

> This corpus contains technical walkthroughs for building AI agent systems.
> Each entry teaches one task where an agent beats a human on speed, consistency, or reliability.
> Every entry contains a before-state, an after-state, and a soul line explaining the advantage.

## Pillars

- Accountability: Proving agent work actually happened
- Memory: Persistent state across agent sessions
- Visibility: Making agent-built sites visible to crawlers
- Agent Teams: Running coordinated multi-agent systems
- Scoping: Killing bad ideas before building
- Operational: Task-specific entries that don't fit a single pillar

## Entries

[empty — first entries arriving soon, see queue/topics.json for the production queue]

## Schema

Tutorial metadata shape: schema/tutorial.schema.json

Every entry has a metadata.json file validated against this schema on every PR via the merge gate (MERGE_GATE.md). Required fields: id, title, slug, pillar, clusterTags, soulLine, beforeState, afterState, status, author, createdAt, lastVerifiedAt.

## Merge gate

Every entry passes seven automated checks before a human reviewer sees it: schema validation, soul line completeness, before/after state completeness, voice violation scan, tool provenance, code block annotations, freshness stamp. Full spec: MERGE_GATE.md.

## Related corpora

- Builder Weekly (the publication): https://thebuilderweekly.com
- The Agent Index (the tool list): https://github.com/thebuilderweekly/the-agent-index

## License

CC BY 4.0 — fork freely, attribute always.