Skip to content

CLAUDE.md

Latest commit

 

History

History
44 lines (33 loc) · 2.23 KB

CLAUDE.md

File metadata and controls

44 lines (33 loc) · 2.23 KB
Raw

Claude Code Instructions

This is a production legacy API for Police Scorecard. It receives public traffic and has external links into API responses, so treat compatibility as the default requirement.

@.claude/rules/annual-scorecard-data.md @.claude/rules/database-migrations.md @.claude/rules/testing.md @.claude/rules/security-and-config.md @.claude/rules/mcp-and-tools.md

Project Shape

  • Runtime: Node 12.12.0, npm 6, Express, Sequelize 5, MySQL 5.7, Elasticsearch 7.5.
  • Entry point: index.js -> app/server.js.
  • API routes: app/api/v1/routes/.
  • Domain logic: app/api/v1/domain/.
  • Sequelize models: app/models/.
  • Migrations: app/migrations/.
  • Tests: Mocha/Chai under test/.
  • Main annual data path: app/data/scorecard.csv and app/api/v1/domain/update.js.

Operating Rules

  • Start with git status --short --branch.
  • Re-read files immediately before editing them.
  • Do not overwrite unrelated user changes.
  • Prefer narrow, additive changes over refactors.
  • Do not upgrade dependencies, runtime versions, build tools, database versions, or lint configuration unless explicitly asked.
  • Do not run deploy scripts, destructive Elasticsearch commands, seed resets, rollback commands, or production API calls unless explicitly asked.
  • Do not add secrets to tracked files. Keep credentials in ignored local config or user-scoped tool configuration.

Common Commands

  • npm test runs the full Mocha test suite.
  • npm run lint runs ESLint.
  • npm run coverage runs tests with nyc.
  • node .claude/skills/annual-scorecard-update/scripts/scorecard-column-audit.js audits the scorecard CSV/header/model/import mapping.
  • docker-compose up --build starts the local Docker stack, but it is slow and touches local containers.

Annual Scorecard Updates

For CSV schema or annual data work, read .claude/skills/annual-scorecard-update/SKILL.md first and run the audit script before editing. The import pipeline is intentionally hard-coded and fragile: every new CSV column must be traced to its owner table, parser, model attribute, migration, import mapping, and API exposure decision.

Do not assume the repo-root scorecard.csv is used by the import path. /update/scorecard validates and imports app/data/scorecard.csv directly.