Vibecode Skribra

You are building a lean indie version of Skribra.

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 =====
# Skribra indie build

## Goal

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

## Scope

One keyword in, one structured long-form article out, published to WordPress · the one destination that doesn't need a platform's permission.

## 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:
- permission to publish anywhere but WordPress · Shopify sits behind App Store review, Webflow behind a marketplace submission, Notion behind a public integration · not code you can write, a queue you wait in
- a bespoke adapter per destination · HTML converted into Notion's block tree with its character and per-request batch limits respected, Webflow field mapping plus a second call to actually publish the item, Shopify find-by-handle upsert
- the feedback loop · Search Console deciding which live posts get refreshed, retitled or relinked, with guards so a page already earning clicks never gets rewritten on a whim
- a pipeline that survives its own failures · every stage retried and resumable, stuck jobs requeued after a deploy kills a worker, and a breaker that trips before the API call so a runaway loop costs zero tokens
- the tuned prompt library and the checks around it · dozens of purpose-built prompts, title coherence and uniqueness gates, and cannibalisation detection so post 20 doesn't quietly compete with post 3

If those capabilities are essential, use Skribra 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 single-article SEO writer that publishes to WordPress. Requirements:

- WordPress only, deliberately: it authenticates with an application password the
  site owner generates alone, while Shopify or Webflow need a platform-approved
  OAuth app; budget weeks of review before one post goes live elsewhere.
- Stack: one Node + TypeScript CLI, no web UI and no accounts, with SQLite via
  better-sqlite3 to remember what's been written. Config in .env.
- Input: a target keyword plus a short site description. Output: one published post.
- Generate in explicit stages, persisting each stage's output to SQLite so a failed run
  resumes instead of restarting: outline · section drafts · intro/conclusion · FAQ · title ·
  meta description · slug.
- Aim for 1500-2500 words of specific prose. Ban the usual tells in the prompt: no 'in today's
  fast-paced world', no 'delve', no section that just restates the one above it.
- Internal links: query the WordPress REST API for existing posts, pick the 3-5 genuinely
  related ones, and link them inline with descriptive anchor text.
- One generated header image, uploaded to the media library and set as featured image.
- Emit JSON-LD Article plus FAQPage, and a real meta description rather than a truncated
  first sentence.
- Publish via the REST API with an application password · draft by default, --publish to go live.
- Log every run's token cost to SQLite and print a per-article total, so the metered cost is
  visible instead of arriving as a surprise.
- Out of scope, and these are the hard parts: keyword research, scheduling, any other CMS,
  and post-publish monitoring or refresh. Pretending a one-shot build covers those would be
  dishonest.

## Required capabilities

- an LLM API key with real per-article spend · this build is metered, not free
- a keyword data source (paid API) or manual keyword picking
- an approved OAuth app per destination beyond WordPress · weeks of someone else's review
- an always-on host with a queue, since one article is dozens of chained calls

## 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 Skribra.

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 =====
# Skribra indie build

## Goal

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

## Scope

One keyword in, one structured long-form article out, published to WordPress · the one destination that doesn't need a platform's permission.

## 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:
- permission to publish anywhere but WordPress · Shopify sits behind App Store review, Webflow behind a marketplace submission, Notion behind a public integration · not code you can write, a queue you wait in
- a bespoke adapter per destination · HTML converted into Notion's block tree with its character and per-request batch limits respected, Webflow field mapping plus a second call to actually publish the item, Shopify find-by-handle upsert
- the feedback loop · Search Console deciding which live posts get refreshed, retitled or relinked, with guards so a page already earning clicks never gets rewritten on a whim
- a pipeline that survives its own failures · every stage retried and resumable, stuck jobs requeued after a deploy kills a worker, and a breaker that trips before the API call so a runaway loop costs zero tokens
- the tuned prompt library and the checks around it · dozens of purpose-built prompts, title coherence and uniqueness gates, and cannibalisation detection so post 20 doesn't quietly compete with post 3

If those capabilities are essential, use Skribra 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 single-article SEO writer that publishes to WordPress. Requirements:

- WordPress only, deliberately: it authenticates with an application password the
  site owner generates alone, while Shopify or Webflow need a platform-approved
  OAuth app; budget weeks of review before one post goes live elsewhere.
- Stack: one Node + TypeScript CLI, no web UI and no accounts, with SQLite via
  better-sqlite3 to remember what's been written. Config in .env.
- Input: a target keyword plus a short site description. Output: one published post.
- Generate in explicit stages, persisting each stage's output to SQLite so a failed run
  resumes instead of restarting: outline · section drafts · intro/conclusion · FAQ · title ·
  meta description · slug.
- Aim for 1500-2500 words of specific prose. Ban the usual tells in the prompt: no 'in today's
  fast-paced world', no 'delve', no section that just restates the one above it.
- Internal links: query the WordPress REST API for existing posts, pick the 3-5 genuinely
  related ones, and link them inline with descriptive anchor text.
- One generated header image, uploaded to the media library and set as featured image.
- Emit JSON-LD Article plus FAQPage, and a real meta description rather than a truncated
  first sentence.
- Publish via the REST API with an application password · draft by default, --publish to go live.
- Log every run's token cost to SQLite and print a per-article total, so the metered cost is
  visible instead of arriving as a surprise.
- Out of scope, and these are the hard parts: keyword research, scheduling, any other CMS,
  and post-publish monitoring or refresh. Pretending a one-shot build covers those would be
  dishonest.

## Required capabilities

- an LLM API key with real per-article spend · this build is metered, not free
- a keyword data source (paid API) or manual keyword picking
- an approved OAuth app per destination beyond WordPress · weeks of someone else's review
- an always-on host with a queue, since one article is dozens of chained calls

## 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 Skribra.

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 =====
# Skribra product brief

## Problem

The writing isn't the hard part · an agent will turn a keyword into a decent long-form article this afternoon, and the prompt below does exactly that. What you can't prompt your way past is permission. Publishing to WordPress is open to anyone with an application password, but every other destination is gatekept by the platform: Shopify wants your app through App Store review, Webflow wants a marketplace submission, Notion wants a public integration, Search Console wants verified OAuth scopes. That's someone else's review queue, measured in weeks, and no model clears it for you. The other half is that content only compounds if something keeps touching it · a generator ships posts and forgets them, and the value here is the unglamorous machinery that comes back to article 14 in week nine because its ranking slipped.

## Product outcome

One keyword in, one structured long-form article out, published to WordPress · the one destination that doesn't need a platform's permission.

## Target user

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

## Required capabilities

- an LLM API key with real per-article spend · this build is metered, not free
- a keyword data source (paid API) or manual keyword picking
- an approved OAuth app per destination beyond WordPress · weeks of someone else's review
- an always-on host with a queue, since one article is dozens of chained calls

## Explicit non-goals for v1

- permission to publish anywhere but WordPress · Shopify sits behind App Store review, Webflow behind a marketplace submission, Notion behind a public integration · not code you can write, a queue you wait in
- a bespoke adapter per destination · HTML converted into Notion's block tree with its character and per-request batch limits respected, Webflow field mapping plus a second call to actually publish the item, Shopify find-by-handle upsert
- the feedback loop · Search Console deciding which live posts get refreshed, retitled or relinked, with guards so a page already earning clicks never gets rewritten on a whim
- a pipeline that survives its own failures · every stage retried and resumable, stuck jobs requeued after a deploy kills a worker, and a breaker that trips before the API call so a runaway loop costs zero tokens
- the tuned prompt library and the checks around it · dozens of purpose-built prompts, title coherence and uniqueness gates, and cannibalisation detection so post 20 doesn't quietly compete with post 3

## 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 single-article SEO writer that publishes to WordPress. Requirements:

- WordPress only, deliberately: it authenticates with an application password the
  site owner generates alone, while Shopify or Webflow need a platform-approved
  OAuth app; budget weeks of review before one post goes live elsewhere.
- Stack: one Node + TypeScript CLI, no web UI and no accounts, with SQLite via
  better-sqlite3 to remember what's been written. Config in .env.
- Input: a target keyword plus a short site description. Output: one published post.
- Generate in explicit stages, persisting each stage's output to SQLite so a failed run
  resumes instead of restarting: outline · section drafts · intro/conclusion · FAQ · title ·
  meta description · slug.
- Aim for 1500-2500 words of specific prose. Ban the usual tells in the prompt: no 'in today's
  fast-paced world', no 'delve', no section that just restates the one above it.
- Internal links: query the WordPress REST API for existing posts, pick the 3-5 genuinely
  related ones, and link them inline with descriptive anchor text.
- One generated header image, uploaded to the media library and set as featured image.
- Emit JSON-LD Article plus FAQPage, and a real meta description rather than a truncated
  first sentence.
- Publish via the REST API with an application password · draft by default, --publish to go live.
- Log every run's token cost to SQLite and print a per-article total, so the metered cost is
  visible instead of arriving as a surprise.
- Out of scope, and these are the hard parts: keyword research, scheduling, any other CMS,
  and post-publish monitoring or refresh. Pretending a one-shot build covers those would be
  dishonest.

## 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 Skribra capabilities as implemented. The v1 non-goals in `PRODUCT.md` remain user-visible limitations until they are deliberately delivered.

# Skribra indie build

## Goal

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

## Scope

One keyword in, one structured long-form article out, published to WordPress · the one destination that doesn't need a platform's permission.

## 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:
- permission to publish anywhere but WordPress · Shopify sits behind App Store review, Webflow behind a marketplace submission, Notion behind a public integration · not code you can write, a queue you wait in
- a bespoke adapter per destination · HTML converted into Notion's block tree with its character and per-request batch limits respected, Webflow field mapping plus a second call to actually publish the item, Shopify find-by-handle upsert
- the feedback loop · Search Console deciding which live posts get refreshed, retitled or relinked, with guards so a page already earning clicks never gets rewritten on a whim
- a pipeline that survives its own failures · every stage retried and resumable, stuck jobs requeued after a deploy kills a worker, and a breaker that trips before the API call so a runaway loop costs zero tokens
- the tuned prompt library and the checks around it · dozens of purpose-built prompts, title coherence and uniqueness gates, and cannibalisation detection so post 20 doesn't quietly compete with post 3

If those capabilities are essential, use Skribra 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 single-article SEO writer that publishes to WordPress. Requirements:

- WordPress only, deliberately: it authenticates with an application password the
  site owner generates alone, while Shopify or Webflow need a platform-approved
  OAuth app; budget weeks of review before one post goes live elsewhere.
- Stack: one Node + TypeScript CLI, no web UI and no accounts, with SQLite via
  better-sqlite3 to remember what's been written. Config in .env.
- Input: a target keyword plus a short site description. Output: one published post.
- Generate in explicit stages, persisting each stage's output to SQLite so a failed run
  resumes instead of restarting: outline · section drafts · intro/conclusion · FAQ · title ·
  meta description · slug.
- Aim for 1500-2500 words of specific prose. Ban the usual tells in the prompt: no 'in today's
  fast-paced world', no 'delve', no section that just restates the one above it.
- Internal links: query the WordPress REST API for existing posts, pick the 3-5 genuinely
  related ones, and link them inline with descriptive anchor text.
- One generated header image, uploaded to the media library and set as featured image.
- Emit JSON-LD Article plus FAQPage, and a real meta description rather than a truncated
  first sentence.
- Publish via the REST API with an application password · draft by default, --publish to go live.
- Log every run's token cost to SQLite and print a per-article total, so the metered cost is
  visible instead of arriving as a surprise.
- Out of scope, and these are the hard parts: keyword research, scheduling, any other CMS,
  and post-publish monitoring or refresh. Pretending a one-shot build covers those would be
  dishonest.

## Required capabilities

- an LLM API key with real per-article spend · this build is metered, not free
- a keyword data source (paid API) or manual keyword picking
- an approved OAuth app per destination beyond WordPress · weeks of someone else's review
- an always-on host with a queue, since one article is dozens of chained calls

## 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.

# Skribra product brief

## Problem

The writing isn't the hard part · an agent will turn a keyword into a decent long-form article this afternoon, and the prompt below does exactly that. What you can't prompt your way past is permission. Publishing to WordPress is open to anyone with an application password, but every other destination is gatekept by the platform: Shopify wants your app through App Store review, Webflow wants a marketplace submission, Notion wants a public integration, Search Console wants verified OAuth scopes. That's someone else's review queue, measured in weeks, and no model clears it for you. The other half is that content only compounds if something keeps touching it · a generator ships posts and forgets them, and the value here is the unglamorous machinery that comes back to article 14 in week nine because its ranking slipped.

## Product outcome

One keyword in, one structured long-form article out, published to WordPress · the one destination that doesn't need a platform's permission.

## Target user

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

## Required capabilities

- an LLM API key with real per-article spend · this build is metered, not free
- a keyword data source (paid API) or manual keyword picking
- an approved OAuth app per destination beyond WordPress · weeks of someone else's review
- an always-on host with a queue, since one article is dozens of chained calls

## Explicit non-goals for v1

- permission to publish anywhere but WordPress · Shopify sits behind App Store review, Webflow behind a marketplace submission, Notion behind a public integration · not code you can write, a queue you wait in
- a bespoke adapter per destination · HTML converted into Notion's block tree with its character and per-request batch limits respected, Webflow field mapping plus a second call to actually publish the item, Shopify find-by-handle upsert
- the feedback loop · Search Console deciding which live posts get refreshed, retitled or relinked, with guards so a page already earning clicks never gets rewritten on a whim
- a pipeline that survives its own failures · every stage retried and resumable, stuck jobs requeued after a deploy kills a worker, and a breaker that trips before the API call so a runaway loop costs zero tokens
- the tuned prompt library and the checks around it · dozens of purpose-built prompts, title coherence and uniqueness gates, and cannibalisation detection so post 20 doesn't quietly compete with post 3

## 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 single-article SEO writer that publishes to WordPress. Requirements:

- WordPress only, deliberately: it authenticates with an application password the
  site owner generates alone, while Shopify or Webflow need a platform-approved
  OAuth app; budget weeks of review before one post goes live elsewhere.
- Stack: one Node + TypeScript CLI, no web UI and no accounts, with SQLite via
  better-sqlite3 to remember what's been written. Config in .env.
- Input: a target keyword plus a short site description. Output: one published post.
- Generate in explicit stages, persisting each stage's output to SQLite so a failed run
  resumes instead of restarting: outline · section drafts · intro/conclusion · FAQ · title ·
  meta description · slug.
- Aim for 1500-2500 words of specific prose. Ban the usual tells in the prompt: no 'in today's
  fast-paced world', no 'delve', no section that just restates the one above it.
- Internal links: query the WordPress REST API for existing posts, pick the 3-5 genuinely
  related ones, and link them inline with descriptive anchor text.
- One generated header image, uploaded to the media library and set as featured image.
- Emit JSON-LD Article plus FAQPage, and a real meta description rather than a truncated
  first sentence.
- Publish via the REST API with an application password · draft by default, --publish to go live.
- Log every run's token cost to SQLite and print a per-article total, so the metered cost is
  visible instead of arriving as a surprise.
- Out of scope, and these are the hard parts: keyword research, scheduling, any other CMS,
  and post-publish monitoring or refresh. Pretending a one-shot build covers those would be
  dishonest.

## 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 Skribra 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