docs/ has grown large (94 files, 29 top-level) with no single entry point.
This index groups every doc by audience so you can find what you need
without reading teach-curriculum.md or any other single file first.
Canonical surface: raw docs/*.md in this directory is the source of
truth. docs-site/v1/*.html is a curated,
smaller subset of the same content for external readers — if the two ever
disagree, trust docs/*.md.
- artifact-format.md — the
eval-report/v1JSON contract, field by field - taxonomy.md — what makes a report "complete"; definitions, examples, checklist
- configuration.md — all config file options
- artifact-format.md
- taxonomy.md
- schema-taxonomy-decisions.md — which concepts become schema fields vs. docs/examples
- judge-axis-rubric-scales.md — LLM-judge axis scoring conventions
- suite-presets.md — built-in agent-quality suite presets
- benchmark-packs.md — versioned safety/tool-routing/groundedness suite bundles
- gates.md — quality gates and CI integration
- publishing.md — GitHub Pages, Azure, custom publish targets
- github-approval-gate-pattern.md — reviewer approvals + commit-status gating
- REPO-HARDENING.md — sign/verify, waivers, heartbeat, bypass tracking
- npm-publishing.md — package release workflow
- teach-curriculum.md — novice path from synthetic dataset to gates/history
- teach-exercises/ — foundational, hands-on exercises (01–11, plus PM-focused pm-01/pm-02)
- teach-labs/ — delivery-stage workflow labs (local dev loop through post-release monitoring, FDE role workflow)
- PROPOSITION-AND-TAXONOMY.md — product north star
- PROPOSITION-REVIEW.md — risks and MVP slice
- PRP.md — original product requirements
- STATUS.md — done vs. remaining, honest completion ledger
- ROADMAP.md — prioritized phases and plan
- ARCHITECTURE.md — system/module structure
- industry-coverage-audit.md — external-source suite/dataset/rubric coverage mapping
- docs-site-stack-decision.md — docs-site stack decision record
- comparison-with-nyc.md — comparison with NYC/Istanbul (code coverage)
- adoption-map.md — candidate existing-runner adoption map
- adoption-feedback-loop.md — weekly page-to-action checks and cadence signals
- adoption-friction-backlog.md — documentation blockers, owner/severity/next action
- adoption-metrics/ — adoption metrics snapshots and manual signal tracker
- community-partnership-log.md — partnership log and outreach stages
- onboarding-runbook.md — new-adopter onboarding steps
- reviews/ — dated review notes
- integrations/README.md — index of "works with" guides
- integrations/promptfoo.md
- integrations/deepeval.md
- integrations/openevals-agentevals.md
- integrations/anthropic-eval-methodology.md
- integrations/langfuse.md
- integrations/wandb-weave.md
- integrations/arize-phoenix.md
- integrations/braintrust.md
- integrations/ragas.md
- integrations/trulens.md
- integrations/patronus.md
- integrations/trace-stacks.md
- integrations/risk-register.md — runtime/version drift, schema drift, cloud coupling risks
- cli-help/ — exact
--helpsnapshots tracked for truth-sync - reporters.md — HTML, text, Markdown, JSON reporters
- examples.md — runnable example index
- case-studies/assistant-ui/README.md — first real external-repo adoption case study