Skip to content

AGENTS.md

Latest commit

 

History

History
53 lines (36 loc) · 2.72 KB

AGENTS.md

File metadata and controls

53 lines (36 loc) · 2.72 KB
Raw

This is NOT the Next.js you know

This version has breaking changes — APIs, conventions, and file structure may all differ from your training data. Read the relevant guide in node_modules/next/dist/docs/ (resolved from this file's directory; in monorepos the next package may not be visible from the repo root) before writing any code. Heed deprecation notices.

This block is written and re-added by next dev — verify at node_modules/next/dist/server/lib/generate-agent-files.js. Removing it from a diff only re-creates the uncommitted change; committing it with your work keeps the tree clean.

Judge Dredd — agent notes

Eval kit: ingest corpora, synthesize and review gold questions, score model answers against rubrics, compare runs, persist history, and optionally ground judgments with hybrid RAG.

Next.js version note

This app uses Next 16.3+. Framework APIs and file conventions may differ from your training data. Prefer the docs bundled with the installed package:

  • node_modules/next/dist/docs/

MCP: root .mcp.json registers next-devtools-mcp. With npm run dev running, agents can attach for errors, routes, and live app state.

Where code lives

Area Path Role
Eval pipeline src/lib/eval/ Judge/complete/parse, campaigns, rubrics, question fixtures, CLIs
Database src/lib/db/ Postgres pool, schema/migrate, evaluate run store, dataset CSV load
RAG src/lib/rag/ Embed, hybrid retrieve, LangGraph query, ground CLI
Connectors src/lib/connectors/ GCP Discovery Engine data stores and Databricks Unity Catalog vector indexes
UI + routes src/app/ App Router pages (/, /batch, /history) and api/* handlers
Components src/components/ Dashboard, playground, verdict, history, batch panels

Leave pure app modules alone unless a change requires them. Human-facing overview: README.md (not a substitute for this file).

Run tests and CLIs

npm test                 # tsx --test; needs local Postgres (CI provides pgvector). No live xAI.
npm run lint
npm run dev              # http://localhost:3000
npm run db:up            # local Postgres (Docker Compose or Homebrew); never AWS
npm run db:migrate
npm run db:smoke
npm run eval:questions   # stub question set CLI
npm run rag:query -- "…"
npm run rag:ground
npm run etl:eval-data    # ETL eval dataset

Env (gitignored .env): DATABASE_URL for local Postgres; XAI_API_KEY (optional XAI_API_KEY2) for live xAI calls. Tests often stub the LLM via EVAL_LLM_STUB or inject fakes—do not require network keys for npm test.

Default local DB URL: postgres://judge:judge@127.0.0.1:5432/synthkit.