@monomit/primer-templates

skill packages and project templates for @monomit/primer

This package contains the template files and domain skill packages used by the primer CLI. It is not intended to be installed directly — install @monomit/primer instead.

Contents

Project templates

templates/cli-tool/ — the base project template rendered during primer init:

cli-tool/
├── AGENTS.md.hbs
├── GETTING_STARTED.md.hbs
├── .cursor/
│   ├── rules/core.mdc.hbs
│   ├── rules/git-discipline.mdc.hbs
│   └── commands/git-*.md.hbs
├── .claude/
│   ├── CLAUDE.md.hbs
│   └── commands/git-*.md.hbs
├── src/index.ts.hbs
├── package.json.hbs
├── tsconfig.json.hbs
├── tsup.config.ts.hbs
├── eslint.config.js.hbs
├── gitignore
└── npmrc

Templates use Mustache (.hbs extension). The scaffold context provides: projectName, packageManager, packageManagerRun, cursorEnabled, claudeEnabled, hasSkills, installedSkillsList, nextStep, finalStep.

AI prompt templates

templates/prompts/generate-agents.hbs — the prompt sent to the LLM during primer init to generate tailored agent role docs. Override in your project by creating .primer/prompt.hbs.

Skill packages

Five domain skill packages are available. Each skill contains:

<skill>/
├── README.md          — when to load this skill
├── rules/<skill>.mdc  — non-negotiable rules (Cursor MDC format)
├── knowledge/         — domain knowledge documents
└── commands/          — step-by-step agent command procedures

skills/database

Knowledge: schema design, migrations, performance, security Commands: design-schema, create-migration, optimize-query, diagnose-performance, audit-schema, audit-security, setup-connection, execute-rollback, analyze-fragmentation, plan-capacity, mask-pii Rules: strict tenant isolation, zero unsafe migrations, parameterized queries only, encryption at rest

skills/auth

Knowledge: authentication patterns, authorization models, secrets management, incident response Commands: setup-auth-provider, configure-access-policy, rotate-secrets, audit-access-control, audit-dependencies, lockdown-auth, revoke-sessions Rules: secrets never in code, MFA for privileged access, short-lived tokens only, auth events always logged

skills/backend

Knowledge: API contracts, resilience patterns, traffic control, observability, agentic backends Commands: design-api-contract, implement-circuit-breaker, configure-rate-limiting, setup-observability, audit-api, scaffold-agentic-backend Rules: validate at boundaries, never swallow errors, all external calls timeout-bounded, structured logging only

skills/frontend

Knowledge: RSC boundary, hydration, FSD architecture, performance Commands: scaffold-architecture, audit-hydration, audit-accessibility, optimize-vitals Rules: strict serialization at RSC boundary, zero randomness during SSR, strict CSP, accessibility as baseline

skills/testing

Knowledge: AI code failure modes, test strategy, mutation testing, contract testing Commands: seed-test-data, generate-mutations, heal-test, diagnose-flakiness Rules: absolute data isolation, no static pauses, quarantine don't ignore, strict pipeline speed limits

Versioning

This package is versioned and released in lockstep with @monomit/primer. The CLI declares a workspace:* dependency on this package — pnpm replaces it with the exact published version during release.

Contributing

To add a new skill package:

1. Create packages/templates/skills/<name>/ with README.md, rules/<name>.mdc, knowledge/, and commands/
2. Add the skill to AVAILABLE_SKILLS in packages/cli/src/lib/skills.ts
3. Add glob patterns to SKILL_GLOBS in packages/cli/src/lib/skills.ts
4. Add installation tests to packages/cli/src/lib/skills.test.ts
5. Open a PR — Copilot will review content accuracy

See the primer monorepo for full development setup.