anthropic-officiallisted

# Anthropic Official + Surface Expert Single lane for **dated official citations** and **which Claude surface fits a goal**. Replaces the old split between thin `anthropic-official` and `anthropic-claude-expert`. ## When to use - Refresh `lab-notes/parity/anthropic-official-posture.md` - Answer: Claude API vs Agent SDK vs `claude -p` vs interactive Claude Code vs playground bridge - Seed policy rows for `lab-notes/parity/claude-parity-matrix.md` - June 15, 2026 metering / Agent SDK credit pool wording (re-fetch help center each time) - OAuth vs API-key routing questions for Alan's policy dialogue ## When not to use - Deep Agents SDK / Messages API doc indexing, changelogs, or project-scoped X watch (**anthropic-platform-expert**) - Implementing runner or bridge code (other agents / default coding) - Live OAuth runs or `localhost:11437` (use **oauth-evidence** after posture is current) - Parity matrix code-audit without policy context (use **parity-archivist**) - Stream-json field inventory (use **observability-scribe**) ## Charter Read [`lab-notes/agents/CHARTER.md`](../../lab-notes/agents/CHARTER.md) before editing. ## Bundled references | File | Purpose | | ---- | ------- | | [sources.md](sources.md) | URL tiers, lookup order, claim labels; ctx7 fallback syntax | | [surfaces.md](surfaces.md) | Surface comparison table and picker | ## How lookups work (plain language) When Alan asks what Anthropic **documents** or which surface to use, **fetch the official page fi