Cruva

Build a local-first TikTok Shop affiliate CRM for one brand.
Use Next.js 15 App Router, TypeScript, Tailwind CSS and SQLite with Drizzle ORM.
The app must run locally with npm install and npm run dev.
Create a dashboard with creators, campaigns, outreach, samples and performance tabs.
Support CSV import for creator profiles, follower counts, category, GMV, sales and content metrics.
Let me score, tag, filter, search and save creators into custom lists.
Add a CRM pipeline with New, Contacted, Replied, Sample Sent, Posted and Won stages.
Track sample status, product, shipping date, content due date and notes per creator.
Add an outreach composer that uses the OpenAI API to draft personalized email or DM copy.
Store the OpenAI key only in .env and provide .env.example.
Add reusable message templates with variables for creator name, product and commission.
Create campaign dashboards for replies, posts, attributed GMV, samples and conversion rates.
Allow CSV export of creators, pipeline records and campaign performance.
Persist all application data in local SQLite with no authentication or cloud database.
Seed the app with realistic demo data so every screen is usable immediately.
Do not scrape TikTok or automate TikTok DMs.
Do not claim access to TikTok Shop APIs, live GMV feeds or Cruva's creator database.
Treat TikTok Shop data as user-supplied CSV imports only.
Use a clean desktop-first interface optimized for managing hundreds of creators.
Include validation, empty states, loading states and clear error messages.
Write a README with setup, CSV column examples, limitations and data backup instructions.
Finish by running typecheck and build, then fix any errors.

# AGENTS.md — Build guide for Cruva

## Project scope
Import creator data, score and segment affiliates, track outreach and samples, generate personalized messages and monitor campaign performance from uploaded CSVs.

Catalogue verdict: no. The CRM and workflow layer are straightforward to rebuild, but Cruva's core value depends on TikTok Shop integrations, a large creator database, live commerce data and outreach infrastructure. A local build can replace internal tracking and message drafting, not the data and platform access that make the product useful at scale.
Use the implementation prompt below to define the deliverable. Complete each phase's acceptance checks before extending the scope.

## Working agreement
- Inspect the repository and its existing instructions before choosing paths, dependencies or commands. Keep one coherent stack and explain changes to the proposed architecture.
- Plan a vertical slice that accepts a real input and produces the useful output described below. Persist only the state the prompt calls for; respect memory-only and upstream-managed workflows. Use fixtures only when they are clearly labelled.
- After scaffolding, document the actual install, development, check and build commands in README and keep them synchronized with the package or project manifest. Do not report commands as successful unless they ran.
- Work in small steps. At handoff, list implemented flows, checks actually performed, remaining blockers, and any credentials or provider setup the owner must supply.
- Do not publish, spend money, contact customers, delete source data or run irreversible migrations without the project owner's authorization.

## Prerequisites
- Node.js 22+
- OpenAI API key
- Creator and performance data exported as CSV

## Stack and architecture
- Use Next.js 15 App Router, TypeScript, Tailwind CSS and SQLite with Drizzle ORM.
- Persist all application data in local SQLite with no authentication or cloud database.
- Interface for this TikTok Shop affiliate outreach + creator CRM workflow: a local web page with input, progress, review, and export views.

## Security and data integrity
- Include validation, empty states, loading states and clear error messages.
- Keep API keys and OAuth secrets in .env; redact them from logs, generated files, and error responses.
- State the practical limit: Cruva's creator database and social intelligence.

## Agent implementation rules
- Project rule, data: Persist all application data in local SQLite with no authentication or cloud database.
- Project rule, behavior: Create a dashboard with creators, campaigns, outreach, samples and performance tabs.
- Project rule, recovery: Include validation, empty states, loading states and clear error messages.

## Optional agent skills and references
No installable skill is listed for this guide; use the project rules and the stack documentation.

Read the linked SKILL.md and its dependencies before adding a skill. Select only the skills matching this project's runtime and task; their documentation does not supply API access, credentials or approval to perform external actions. Pin the reviewed revision where the tool supports it. Follow the chosen agent's documented project-level installation mechanism.

## Distribution ideas
These are optional planning notes. Obtain the owner's approval before publishing or contacting anyone.
- Demonstrate this TikTok Shop affiliate outreach + creator CRM operation with sample or user-owned data: Create a dashboard with creators, campaigns, outreach, samples and performance tabs.
- Show a concrete result or export: Allow CSV export of creators, pipeline records and campaign performance.
- State the practical limit: Cruva's creator database and social intelligence.

## Engineering roadmap
1. Phase 1, architecture and data: Use Next.js 15 App Router, TypeScript, Tailwind CSS and SQLite with Drizzle ORM. Persist all application data in local SQLite with no authentication or cloud database.
2. Phase 2, implement: Create a dashboard with creators, campaigns, outreach, samples and performance tabs.
3. Phase 3, implement: Support CSV import for creator profiles, follower counts, category, GMV, sales and content metrics.
4. Phase 4, review and output: Add a CRM pipeline with New, Contacted, Replied, Sample Sent, Posted and Won stages. Allow CSV export of creators, pipeline records and campaign performance.
5. Phase 5, recovery and acceptance: Include validation, empty states, loading states and clear error messages. Verify this invariant with a saved fixture: A repeated payment webhook cannot deliver twice; a refund must revoke the relevant access without deleting the audit trail. State the practical limit: Cruva's creator database and social intelligence.

## Paid-product capabilities outside this build
- Cruva's creator database and social intelligence
- Native TikTok Shop integrations and live GMV data
- Automated TikTok DMs and target collaboration outreach
- Built-in sample request and affiliate automation at scale
- Managed data enrichment and deliverability infrastructure

## Implementation prompt
Build a local-first TikTok Shop affiliate CRM for one brand.
Use Next.js 15 App Router, TypeScript, Tailwind CSS and SQLite with Drizzle ORM.
The app must run locally with npm install and npm run dev.
Create a dashboard with creators, campaigns, outreach, samples and performance tabs.
Support CSV import for creator profiles, follower counts, category, GMV, sales and content metrics.
Let me score, tag, filter, search and save creators into custom lists.
Add a CRM pipeline with New, Contacted, Replied, Sample Sent, Posted and Won stages.
Track sample status, product, shipping date, content due date and notes per creator.
Add an outreach composer that uses the OpenAI API to draft personalized email or DM copy.
Store the OpenAI key only in .env and provide .env.example.
Add reusable message templates with variables for creator name, product and commission.
Create campaign dashboards for replies, posts, attributed GMV, samples and conversion rates.
Allow CSV export of creators, pipeline records and campaign performance.
Persist all application data in local SQLite with no authentication or cloud database.
Seed the app with realistic demo data so every screen is usable immediately.
Do not scrape TikTok or automate TikTok DMs.
Do not claim access to TikTok Shop APIs, live GMV feeds or Cruva's creator database.
Treat TikTok Shop data as user-supplied CSV imports only.
Use a clean desktop-first interface optimized for managing hundreds of creators.
Include validation, empty states, loading states and clear error messages.
Write a README with setup, CSV column examples, limitations and data backup instructions.
Finish by running typecheck and build, then fix any errors.

## Completion evidence
Demonstrate the prompt's acceptance scenarios against the scoped workflow. Include setup from a clean checkout and failure recovery. Check persistence across restart and export/restore only for the state the prompt says to store; for memory-only tools, confirm that temporary content is discarded as specified. Record actual results and remaining limitations. A detailed plan alone does not establish a working replacement.