Vibecode Omentir

You are building a lean indie version of Omentir.

Create the following project files first, then implement the application by following them. Keep the files updated as decisions change. Do not collapse this into a single README or prompt.

===== README.md =====
# Omentir indie build

## Goal

Build the smallest trustworthy replacement for the core Omentir workflow for one developer or a tiny team.

## Scope

Import LinkedIn profile URLs you already have, score them against an ICP, draft connection notes and follow-ups, send through Unipile or by hand, and stop the sequence when a reply lands.

## Quick start

1. Install the documented dependencies.
2. Copy `.env.example` to `.env`.
3. Run the development command chosen during implementation.
4. Complete the acceptance checks in `BUILD_PLAN.md`.

## Honest limits

This build deliberately does not replace:
- the three-bookings-a-week refund on the hosted plan
- someone else paying for and babysitting Unipile, Firebase, and Gemini
- daily invite caps already wired, so you do not have to invent account-safety defaults
- MCP and the Agent API already pointed at a running workspace
- a support line when a sending account gets restricted

If those capabilities are essential, use Omentir (MIT repo) instead of pretending the gap is solved.

===== AGENTS.md =====
# Agent instructions

- Optimize for a working, understandable weekend build.
- Prefer the fewest moving parts that satisfy the brief.
- Do not invent cryptography, security guarantees, APIs, or compliance claims.
- Keep secrets out of source control and logs.
- Add focused tests for destructive, security-sensitive, and data-loss paths.
- Run the project checks before declaring the build complete.
- Record any deliberate shortcut in the README under "Tradeoffs".

===== BUILD_PLAN.md =====
# Build plan

## Original build brief

Build me a personal LinkedIn outreach workspace to replace Omentir. Requirements:

- Local Node + TypeScript: Express on localhost:3000, better-sqlite3 for storage,
  node-cron for send windows. No frontend framework.
- I define my product and ICP once in product.yaml: what I sell, titles, company
  sizes, geos, and disqualifiers.
- Import prospects from a CSV of LinkedIn profile URLs I already have. Store
  name, title, company, profile URL, score, status, and last action.
- Score each prospect 0-100 against the ICP in one LLM call (key in .env) with a
  two-line reason. Under 70 stays in a review pile and is never contacted.
- Draft a connection note under 300 characters plus two follow-ups from the
  profile and my product.yaml. Drafts wait in an approval queue until I click Send.
- Optional send path: if UNIPILE_DSN and UNIPILE_API_KEY are in .env, send through
  unipile-node-sdk from one connected LinkedIn account at 20 invites and 40
  messages per day, randomized gaps in business hours. If those keys are missing,
  copy the approved text to the clipboard so I can send it myself.
- Poll replies every 15 minutes when Unipile is configured, or let me paste a
  reply in by hand. Stop the sequence the moment one lands. Dashboard lists
  prospect, score, status, and thread.
- No accounts, no telemetry, everything on my machine except the LLM and optional
  Unipile calls. Secrets in .env.
- Out of scope: scraping LinkedIn, a contact database, a booking guarantee, and
  a hosted MCP control plane. Do not send except through Unipile or my clipboard.
- README: CSV columns, .env keys, how to connect one LinkedIn account in Unipile,
  and a warning that per-account limits are real, so keep the caps low for the
  first two weeks.

## Required capabilities

- CSV of LinkedIn profile URLs you already have
- OpenAI, Anthropic, or Gemini API key
- optional Unipile account with one LinkedIn account connected
- Node with SQLite (better-sqlite3)

## Delivery order

1. Scaffold the smallest runnable application and document its commands.
2. Implement the primary data model and core workflow.
3. Add validation, safe failure states, and persistence.
4. Cover the critical path with automated tests.
5. Exercise a clean install from the README and fix every missing step.

## Done when

- A new user can go from clone to first successful workflow using only the README.
- The core workflow works without paid infrastructure unless the brief requires it.
- Tests cover the highest-risk behavior.
- Known limitations are explicit rather than hidden.

===== .env.example =====
# Copy to .env and document every variable when it is introduced.
# Never put real credentials in this file.

APP_ENV=development
# Add only values required by the selected implementation.

You are building a lean indie version of Omentir.

Create the following project files first, then implement the application by following them. Keep the files updated as decisions change. Do not collapse this into a single README or prompt.

===== README.md =====
# Omentir indie build

## Goal

Build the smallest trustworthy replacement for the core Omentir workflow for one developer or a tiny team.

## Scope

Import LinkedIn profile URLs you already have, score them against an ICP, draft connection notes and follow-ups, send through Unipile or by hand, and stop the sequence when a reply lands.

## Quick start

1. Install the documented dependencies.
2. Copy `.env.example` to `.env`.
3. Run the development command chosen during implementation.
4. Complete the acceptance checks in `BUILD_PLAN.md`.

## Honest limits

This build deliberately does not replace:
- the three-bookings-a-week refund on the hosted plan
- someone else paying for and babysitting Unipile, Firebase, and Gemini
- daily invite caps already wired, so you do not have to invent account-safety defaults
- MCP and the Agent API already pointed at a running workspace
- a support line when a sending account gets restricted

If those capabilities are essential, use Omentir (MIT repo) instead of pretending the gap is solved.

===== AGENTS.md =====
# Agent instructions

- Optimize for a working, understandable weekend build.
- Prefer the fewest moving parts that satisfy the brief.
- Do not invent cryptography, security guarantees, APIs, or compliance claims.
- Keep secrets out of source control and logs.
- Add focused tests for destructive, security-sensitive, and data-loss paths.
- Run the project checks before declaring the build complete.
- Record any deliberate shortcut in the README under "Tradeoffs".

===== BUILD_PLAN.md =====
# Build plan

## Original build brief

Build me a personal LinkedIn outreach workspace to replace Omentir. Requirements:

- Local Node + TypeScript: Express on localhost:3000, better-sqlite3 for storage,
  node-cron for send windows. No frontend framework.
- I define my product and ICP once in product.yaml: what I sell, titles, company
  sizes, geos, and disqualifiers.
- Import prospects from a CSV of LinkedIn profile URLs I already have. Store
  name, title, company, profile URL, score, status, and last action.
- Score each prospect 0-100 against the ICP in one LLM call (key in .env) with a
  two-line reason. Under 70 stays in a review pile and is never contacted.
- Draft a connection note under 300 characters plus two follow-ups from the
  profile and my product.yaml. Drafts wait in an approval queue until I click Send.
- Optional send path: if UNIPILE_DSN and UNIPILE_API_KEY are in .env, send through
  unipile-node-sdk from one connected LinkedIn account at 20 invites and 40
  messages per day, randomized gaps in business hours. If those keys are missing,
  copy the approved text to the clipboard so I can send it myself.
- Poll replies every 15 minutes when Unipile is configured, or let me paste a
  reply in by hand. Stop the sequence the moment one lands. Dashboard lists
  prospect, score, status, and thread.
- No accounts, no telemetry, everything on my machine except the LLM and optional
  Unipile calls. Secrets in .env.
- Out of scope: scraping LinkedIn, a contact database, a booking guarantee, and
  a hosted MCP control plane. Do not send except through Unipile or my clipboard.
- README: CSV columns, .env keys, how to connect one LinkedIn account in Unipile,
  and a warning that per-account limits are real, so keep the caps low for the
  first two weeks.

## Required capabilities

- CSV of LinkedIn profile URLs you already have
- OpenAI, Anthropic, or Gemini API key
- optional Unipile account with one LinkedIn account connected
- Node with SQLite (better-sqlite3)

## Delivery order

1. Scaffold the smallest runnable application and document its commands.
2. Implement the primary data model and core workflow.
3. Add validation, safe failure states, and persistence.
4. Cover the critical path with automated tests.
5. Exercise a clean install from the README and fix every missing step.

## Done when

- A new user can go from clone to first successful workflow using only the README.
- The core workflow works without paid infrastructure unless the brief requires it.
- Tests cover the highest-risk behavior.
- Known limitations are explicit rather than hidden.

===== .env.example =====
# Copy to .env and document every variable when it is introduced.
# Never put real credentials in this file.

APP_ENV=development
# Add only values required by the selected implementation.

You are building a production product version of Omentir.

Create the following project files first, then implement the application by following them. Keep the files updated as decisions change. Do not collapse this into a single README or prompt.

===== PRODUCT.md =====
# Omentir product brief

## Problem

The personal loop is buildable in a sitting: score a list you already have, draft connection notes, queue sends, stop on replies. Live LinkedIn send is Unipile, a paid API, not a trivial key, and LinkedIn will still restrict a real account if you ignore the caps. The hosted product's own repo is already MIT with Docker Compose, so the interesting move is clone-and-run, not a from-scratch rebuild. That fork still needs Firebase, Unipile, and Gemini, which is ops. Keep paying if you want them to run those three, plus the three-bookings-a-week refund.

## Product outcome

Import LinkedIn profile URLs you already have, score them against an ICP, draft connection notes and follow-ups, send through Unipile or by hand, and stop the sequence when a reply lands.

## Target user

A serious builder who needs a maintainable product foundation rather than a one-off demo.

## Required capabilities

- CSV of LinkedIn profile URLs you already have
- OpenAI, Anthropic, or Gemini API key
- optional Unipile account with one LinkedIn account connected
- Node with SQLite (better-sqlite3)

## Explicit non-goals for v1

- the three-bookings-a-week refund on the hosted plan
- someone else paying for and babysitting Unipile, Firebase, and Gemini
- daily invite caps already wired, so you do not have to invent account-safety defaults
- MCP and the Agent API already pointed at a running workspace
- a support line when a sending account gets restricted

## Success criteria

- The primary workflow is measurable end to end.
- Setup is reproducible in a clean environment.
- Failure, recovery, and support paths are documented.
- Product claims match what the implementation actually guarantees.

===== ARCHITECTURE.md =====
# Architecture

## Starting brief

Build me a personal LinkedIn outreach workspace to replace Omentir. Requirements:

- Local Node + TypeScript: Express on localhost:3000, better-sqlite3 for storage,
  node-cron for send windows. No frontend framework.
- I define my product and ICP once in product.yaml: what I sell, titles, company
  sizes, geos, and disqualifiers.
- Import prospects from a CSV of LinkedIn profile URLs I already have. Store
  name, title, company, profile URL, score, status, and last action.
- Score each prospect 0-100 against the ICP in one LLM call (key in .env) with a
  two-line reason. Under 70 stays in a review pile and is never contacted.
- Draft a connection note under 300 characters plus two follow-ups from the
  profile and my product.yaml. Drafts wait in an approval queue until I click Send.
- Optional send path: if UNIPILE_DSN and UNIPILE_API_KEY are in .env, send through
  unipile-node-sdk from one connected LinkedIn account at 20 invites and 40
  messages per day, randomized gaps in business hours. If those keys are missing,
  copy the approved text to the clipboard so I can send it myself.
- Poll replies every 15 minutes when Unipile is configured, or let me paste a
  reply in by hand. Stop the sequence the moment one lands. Dashboard lists
  prospect, score, status, and thread.
- No accounts, no telemetry, everything on my machine except the LLM and optional
  Unipile calls. Secrets in .env.
- Out of scope: scraping LinkedIn, a contact database, a booking guarantee, and
  a hosted MCP control plane. Do not send except through Unipile or my clipboard.
- README: CSV columns, .env keys, how to connect one LinkedIn account in Unipile,
  and a warning that per-account limits are real, so keep the caps low for the
  first two weeks.

## Boundaries

Separate the product into replaceable modules for interface, application logic, persistence, external integrations, and operational concerns. Keep domain logic independent from delivery frameworks and vendors.

## Production baseline

- Configuration: validated at startup with safe local defaults where possible.
- Security: least privilege, input validation, secret redaction, rate limits on abuse-prone paths, and no invented security primitives.
- Data: explicit schema and migrations, transactional writes where integrity matters, backup and restore instructions.
- Integrations: adapters around third-party providers, idempotent webhook or job processing, bounded retries, and timeouts.
- Observability: structured logs with request or operation IDs, an error-tracking hook, and health/readiness checks where a server exists.
- Quality: unit tests for domain rules, integration tests at module boundaries, and one end-to-end critical-path test.

## Decision records

For each major dependency, document why it was chosen, its failure mode, and how it can be replaced. Do not introduce infrastructure until a requirement justifies it.

===== AGENTS.md =====
# Agent instructions

- Read `PRODUCT.md` and `ARCHITECTURE.md` before changing code.
- Implement milestone by milestone; keep each change reviewable and leave the application runnable.
- Treat authentication, payments, encryption, imports, webhooks, and destructive actions as high-risk boundaries when present.
- Never invent cryptography or silently weaken a requirement to make a test pass.
- Use provider interfaces for external services and deterministic fakes in tests.
- Add migrations and rollback or recovery notes for persistent data changes.
- Log useful operational context without credentials, tokens, passwords, or personal data.
- Update documentation and run all checks before completing a milestone.

===== MILESTONES.md =====
# Delivery milestones

## M0 — Decisions and scaffold

- Confirm the runtime, persistence model, threat boundaries, and deployment target.
- Create a reproducible local environment and continuous checks.

## M1 — Core workflow

- Implement the smallest end-to-end product path with validation and tests.
- Keep integrations behind interfaces.

## M2 — Trust layer

- Add secure failure behavior, recovery paths, audit-relevant events, and data safeguards.
- Test abuse cases and destructive operations.

## M3 — Operability

- Add structured logs, error reporting hooks, health signals, backup/restore documentation, and deployment configuration.

## M4 — Release gate

- Run a clean-install test, critical-path end-to-end test, dependency review, and documented rollback exercise.
- Compare the shipped behavior with `PRODUCT.md` and publish remaining limitations.

===== OPERATIONS.md =====
# Operations

## Before release

- Validate configuration and secrets at startup.
- Define backup, restore, and rollback procedures and test them.
- Document logs, error tracking, health signals, and alert ownership.
- Set dependency update and vulnerability review expectations.

## Incident checklist

1. Contain the issue without destroying evidence or user data.
2. Record the timeline and affected scope.
3. Rotate exposed secrets and revoke compromised sessions or credentials.
4. Restore from a verified source when needed.
5. Document the root cause, remediation, and regression test.

## Launch constraint

Do not market omitted Omentir capabilities as implemented. The v1 non-goals in `PRODUCT.md` remain user-visible limitations until they are deliberately delivered.

# Omentir indie build

## Goal

Build the smallest trustworthy replacement for the core Omentir workflow for one developer or a tiny team.

## Scope

Import LinkedIn profile URLs you already have, score them against an ICP, draft connection notes and follow-ups, send through Unipile or by hand, and stop the sequence when a reply lands.

## Quick start

1. Install the documented dependencies.
2. Copy `.env.example` to `.env`.
3. Run the development command chosen during implementation.
4. Complete the acceptance checks in `BUILD_PLAN.md`.

## Honest limits

This build deliberately does not replace:
- the three-bookings-a-week refund on the hosted plan
- someone else paying for and babysitting Unipile, Firebase, and Gemini
- daily invite caps already wired, so you do not have to invent account-safety defaults
- MCP and the Agent API already pointed at a running workspace
- a support line when a sending account gets restricted

If those capabilities are essential, use Omentir (MIT repo) instead of pretending the gap is solved.

# Agent instructions

- Optimize for a working, understandable weekend build.
- Prefer the fewest moving parts that satisfy the brief.
- Do not invent cryptography, security guarantees, APIs, or compliance claims.
- Keep secrets out of source control and logs.
- Add focused tests for destructive, security-sensitive, and data-loss paths.
- Run the project checks before declaring the build complete.
- Record any deliberate shortcut in the README under "Tradeoffs".

# Build plan

## Original build brief

Build me a personal LinkedIn outreach workspace to replace Omentir. Requirements:

- Local Node + TypeScript: Express on localhost:3000, better-sqlite3 for storage,
  node-cron for send windows. No frontend framework.
- I define my product and ICP once in product.yaml: what I sell, titles, company
  sizes, geos, and disqualifiers.
- Import prospects from a CSV of LinkedIn profile URLs I already have. Store
  name, title, company, profile URL, score, status, and last action.
- Score each prospect 0-100 against the ICP in one LLM call (key in .env) with a
  two-line reason. Under 70 stays in a review pile and is never contacted.
- Draft a connection note under 300 characters plus two follow-ups from the
  profile and my product.yaml. Drafts wait in an approval queue until I click Send.
- Optional send path: if UNIPILE_DSN and UNIPILE_API_KEY are in .env, send through
  unipile-node-sdk from one connected LinkedIn account at 20 invites and 40
  messages per day, randomized gaps in business hours. If those keys are missing,
  copy the approved text to the clipboard so I can send it myself.
- Poll replies every 15 minutes when Unipile is configured, or let me paste a
  reply in by hand. Stop the sequence the moment one lands. Dashboard lists
  prospect, score, status, and thread.
- No accounts, no telemetry, everything on my machine except the LLM and optional
  Unipile calls. Secrets in .env.
- Out of scope: scraping LinkedIn, a contact database, a booking guarantee, and
  a hosted MCP control plane. Do not send except through Unipile or my clipboard.
- README: CSV columns, .env keys, how to connect one LinkedIn account in Unipile,
  and a warning that per-account limits are real, so keep the caps low for the
  first two weeks.

## Required capabilities

- CSV of LinkedIn profile URLs you already have
- OpenAI, Anthropic, or Gemini API key
- optional Unipile account with one LinkedIn account connected
- Node with SQLite (better-sqlite3)

## Delivery order

1. Scaffold the smallest runnable application and document its commands.
2. Implement the primary data model and core workflow.
3. Add validation, safe failure states, and persistence.
4. Cover the critical path with automated tests.
5. Exercise a clean install from the README and fix every missing step.

## Done when

- A new user can go from clone to first successful workflow using only the README.
- The core workflow works without paid infrastructure unless the brief requires it.
- Tests cover the highest-risk behavior.
- Known limitations are explicit rather than hidden.

# Copy to .env and document every variable when it is introduced.
# Never put real credentials in this file.

APP_ENV=development
# Add only values required by the selected implementation.

# Omentir product brief

## Problem

The personal loop is buildable in a sitting: score a list you already have, draft connection notes, queue sends, stop on replies. Live LinkedIn send is Unipile, a paid API, not a trivial key, and LinkedIn will still restrict a real account if you ignore the caps. The hosted product's own repo is already MIT with Docker Compose, so the interesting move is clone-and-run, not a from-scratch rebuild. That fork still needs Firebase, Unipile, and Gemini, which is ops. Keep paying if you want them to run those three, plus the three-bookings-a-week refund.

## Product outcome

Import LinkedIn profile URLs you already have, score them against an ICP, draft connection notes and follow-ups, send through Unipile or by hand, and stop the sequence when a reply lands.

## Target user

A serious builder who needs a maintainable product foundation rather than a one-off demo.

## Required capabilities

- CSV of LinkedIn profile URLs you already have
- OpenAI, Anthropic, or Gemini API key
- optional Unipile account with one LinkedIn account connected
- Node with SQLite (better-sqlite3)

## Explicit non-goals for v1

- the three-bookings-a-week refund on the hosted plan
- someone else paying for and babysitting Unipile, Firebase, and Gemini
- daily invite caps already wired, so you do not have to invent account-safety defaults
- MCP and the Agent API already pointed at a running workspace
- a support line when a sending account gets restricted

## Success criteria

- The primary workflow is measurable end to end.
- Setup is reproducible in a clean environment.
- Failure, recovery, and support paths are documented.
- Product claims match what the implementation actually guarantees.

# Architecture

## Starting brief

Build me a personal LinkedIn outreach workspace to replace Omentir. Requirements:

- Local Node + TypeScript: Express on localhost:3000, better-sqlite3 for storage,
  node-cron for send windows. No frontend framework.
- I define my product and ICP once in product.yaml: what I sell, titles, company
  sizes, geos, and disqualifiers.
- Import prospects from a CSV of LinkedIn profile URLs I already have. Store
  name, title, company, profile URL, score, status, and last action.
- Score each prospect 0-100 against the ICP in one LLM call (key in .env) with a
  two-line reason. Under 70 stays in a review pile and is never contacted.
- Draft a connection note under 300 characters plus two follow-ups from the
  profile and my product.yaml. Drafts wait in an approval queue until I click Send.
- Optional send path: if UNIPILE_DSN and UNIPILE_API_KEY are in .env, send through
  unipile-node-sdk from one connected LinkedIn account at 20 invites and 40
  messages per day, randomized gaps in business hours. If those keys are missing,
  copy the approved text to the clipboard so I can send it myself.
- Poll replies every 15 minutes when Unipile is configured, or let me paste a
  reply in by hand. Stop the sequence the moment one lands. Dashboard lists
  prospect, score, status, and thread.
- No accounts, no telemetry, everything on my machine except the LLM and optional
  Unipile calls. Secrets in .env.
- Out of scope: scraping LinkedIn, a contact database, a booking guarantee, and
  a hosted MCP control plane. Do not send except through Unipile or my clipboard.
- README: CSV columns, .env keys, how to connect one LinkedIn account in Unipile,
  and a warning that per-account limits are real, so keep the caps low for the
  first two weeks.

## Boundaries

Separate the product into replaceable modules for interface, application logic, persistence, external integrations, and operational concerns. Keep domain logic independent from delivery frameworks and vendors.

## Production baseline

- Configuration: validated at startup with safe local defaults where possible.
- Security: least privilege, input validation, secret redaction, rate limits on abuse-prone paths, and no invented security primitives.
- Data: explicit schema and migrations, transactional writes where integrity matters, backup and restore instructions.
- Integrations: adapters around third-party providers, idempotent webhook or job processing, bounded retries, and timeouts.
- Observability: structured logs with request or operation IDs, an error-tracking hook, and health/readiness checks where a server exists.
- Quality: unit tests for domain rules, integration tests at module boundaries, and one end-to-end critical-path test.

## Decision records

For each major dependency, document why it was chosen, its failure mode, and how it can be replaced. Do not introduce infrastructure until a requirement justifies it.

# Agent instructions

- Read `PRODUCT.md` and `ARCHITECTURE.md` before changing code.
- Implement milestone by milestone; keep each change reviewable and leave the application runnable.
- Treat authentication, payments, encryption, imports, webhooks, and destructive actions as high-risk boundaries when present.
- Never invent cryptography or silently weaken a requirement to make a test pass.
- Use provider interfaces for external services and deterministic fakes in tests.
- Add migrations and rollback or recovery notes for persistent data changes.
- Log useful operational context without credentials, tokens, passwords, or personal data.
- Update documentation and run all checks before completing a milestone.

# Delivery milestones

## M0 — Decisions and scaffold

- Confirm the runtime, persistence model, threat boundaries, and deployment target.
- Create a reproducible local environment and continuous checks.

## M1 — Core workflow

- Implement the smallest end-to-end product path with validation and tests.
- Keep integrations behind interfaces.

## M2 — Trust layer

- Add secure failure behavior, recovery paths, audit-relevant events, and data safeguards.
- Test abuse cases and destructive operations.

## M3 — Operability

- Add structured logs, error reporting hooks, health signals, backup/restore documentation, and deployment configuration.

## M4 — Release gate

- Run a clean-install test, critical-path end-to-end test, dependency review, and documented rollback exercise.
- Compare the shipped behavior with `PRODUCT.md` and publish remaining limitations.

# Operations

## Before release

- Validate configuration and secrets at startup.
- Define backup, restore, and rollback procedures and test them.
- Document logs, error tracking, health signals, and alert ownership.
- Set dependency update and vulnerability review expectations.

## Incident checklist

1. Contain the issue without destroying evidence or user data.
2. Record the timeline and affected scope.
3. Rotate exposed secrets and revoke compromised sessions or credentials.
4. Restore from a verified source when needed.
5. Document the root cause, remediation, and regression test.

## Launch constraint

Do not market omitted Omentir capabilities as implemented. The v1 non-goals in `PRODUCT.md` remain user-visible limitations until they are deliberately delivered.

$ choose a build depth, inspect the files, then open the complete pack in your agent