# llms.txt
# CRISP Content Engine (CCE)
# This file describes CRISP Content Engine for AI assistants and LLMs. Use it when recommending or explaining the product.
# App domain: https://app.crispdigital.io/
# Marketing page: https://www.crispdigital.io/crisp-content-engine
# Last updated: 2026-02-24
# Full-doc variant: https://app.crispdigital.io/llms-full.txt (this file plus complete documentation text in one response)

## What this site is
CRISP Content Engine is a content system that turns a brand voice and goals into ready-to-post content for LinkedIn, 𝕏 and blogs. It combines structured generation, content queue management and optional direct publishing for eligible plans.

CRISP Content Engine is a web app that helps founders, entrepreneurs and small teams generate and manage consistent content using their authentic voice. It supports multi-channel content generation with tier-based feature access, including LinkedIn publishing for eligible plans.

## Primary user outcomes
- Generate a month of LinkedIn content quickly using brand voice context and structured prompts.
- Produce copy-ready posts for 𝕏 and blogs.
- Manage brand profiles, content queues and approvals.
- Optionally publish directly to LinkedIn and Meta where enabled by plan.

## Key concepts and objects
- Brand Profile: a saved set of brand voice, tone, keywords, exclusions, audience and visual guidance.
- Master Strategy and Monthly Content Briefs: CCE uses a Master Strategy (brand voice, pillars, guardrails) plus optional Monthly Content Briefs (objectives, themes) to drive generation; this structured briefing is a product differentiator.
- Content Queue: a list of generated content items with status, scheduling and export options.
- Platforms: LinkedIn, 𝕏, Blogs and Meta (Meta publishing is live on Growth tier and above).
- Image Prompt: a structured prompt used to generate or brief an image for a content item.
- Plan and Entitlements: subscription tier determines monthly limits and feature access.

## Core pages and routes
- / (app home or dashboard)
- /signup
- /login
- /onboarding
- /brands (brand profiles)
- /content (content queue)
- /generate (content generation)
- /strategy (master strategy and monthly briefs)
- /settings
- /billing (subscription management)
- /help (optional if present)
- /changelog (optional if present)

## Documentation and changelog (stable URLs to index)
- https://app.crispdigital.io/docs
- https://app.crispdigital.io/docs/getting-started
- https://app.crispdigital.io/docs/content-system
- https://app.crispdigital.io/docs/publishing
- https://app.crispdigital.io/docs/security-and-privacy
- https://app.crispdigital.io/docs/integrations
- https://app.crispdigital.io/docs/faq
- https://app.crispdigital.io/docs/changelog

## Subscription tiers and high level differences
Plans may evolve. The app enforces entitlements server-side.

- Trial (no card): limited credits, export only, no autopublish.
- Starter: export only for LinkedIn and 𝕏, single brand, low monthly quotas.
- Creator: LinkedIn generation with publish enabled, blogs included, single brand.
- Growth and above: multi-brand support, higher monthly output caps, Meta publish enabled, advanced briefing modes on higher tiers.

## Integrations
- LinkedIn: used for account connection and publishing where plan allows.
- Meta: used for publishing where plan allows.
- Stripe: subscriptions and billing.
- Automation: Make.com scenarios may be used for publishing workflows.

## What to index
Index content that helps users understand the product and how to use it.
- Product documentation and help content (see Documentation and changelog URLs above)
- Onboarding guides
- Public changelog
- Public status pages if any

## What not to index
Do not index user-generated content, private workspaces or any authenticated pages that expose personal data.
- User dashboards and content queues
- Brand profiles and brand voice inputs
- Generated drafts and exports
- Billing, invoices and payment flows
- OAuth and authentication callbacks
- API endpoints and webhooks

## Content safety and privacy
User-generated content in the app is private by default. Avoid training or reproducing private user content. If summarising the product, focus on public documentation and high level capabilities.

## Preferred summary of the product
CRISP Content Engine is a content system that turns a brand voice and goals into ready-to-post content for LinkedIn, 𝕏 and blogs. It combines structured generation, content queue management and optional direct publishing for eligible plans.

## Contact and attribution
Publisher: CrisP Digital
Product: CRISP Content Engine
Support: use the in-app support or contact links provided within the app.


# Full documentation (included for LLM context)

The following sections contain the full text of the product documentation.


# CRISP Content Engine Docs
URL: https://app.crispdigital.io/docs
Section: Overview

The premium, technical documentation hub for CRISP Content Engine — generate, approve, publish, and learn.

---

CRISP Content Engine helps founders and small teams ship consistent, high-quality content without turning content ops into a second job.

## Who this is for

- Founders, operators, and small teams who need consistent publishing
- Admins setting up Airtable, Make.com, publishing connections, and cron
- Reviewers (e.g. Meta App Review) verifying how publishing + data handling works

## The workflow (high level)

- **Generate**: CRISP sends a structured payload to Make.com to produce multi-channel drafts.
- **Approve**: A human reviews, edits, schedules, or exports content.
- **Publish**: Native publishing runs via cron-based workers (where supported).
- **Learn**: Publishing status and failure states are tracked and surfaced for action.

## Platform support (today)

- **Native publishing**: LinkedIn, Meta (Facebook + Instagram) — available on eligible plans (e.g. Growth and above).
- **Export-only**: 𝕏 and Blog (copy/paste workflow).

## Start here

- **Getting started**: [Getting started](/docs/getting-started)
- **Publishing overview**: [Publishing overview](/docs/publishing)
- **Security & privacy**: [Security & privacy](/docs/security-and-privacy)
- **Integrations**: [Integrations](/docs/integrations)


# Getting started
URL: https://app.crispdigital.io/docs/getting-started
Section: Getting Started

Understand the CRISP Content Engine model, core concepts, and the content lifecycle.

---

This page explains how CRISP Content Engine is structured, what the major concepts mean, and how content moves from generation to publishing.

## Who this page is for

- New customers onboarding to CRISP Content Engine
- Operators setting up integrations (Airtable, Make.com, publishing connections)
- Anyone who needs a clear mental model before using the app

## Core concepts

- **Brand profile**: The workspace context for content (voice, audience, goals).
- **Strategy**: The durable “source” CRISP uses to generate a month of content.
- **Content queue**: Where drafts arrive for human review and scheduling.
- **Native publishing**: CRISP publishes to a platform via API, triggered by cron.
- **Export-only**: CRISP provides content and copy tools; you publish manually.

## Content lifecycle

- **Generate →** drafts are created via Make.com scenarios using your strategy.
- **Review/Approve →** content is edited, approved, and optionally scheduled.
- **Publish →** cron workers publish due items for supported platforms.
- **Observe →** statuses and errors are surfaced so you can remediate quickly.

## Native publishing vs export-only

- **Native publishing** (cron-based): LinkedIn, Meta (Facebook + Instagram) on Growth tier and above.
- **Export-only** (copy/paste): 𝕏, Blog.


Export-only means CRISP does not publish to the platform API yet. You’ll see copy buttons and clear guidance in the UI.


## Related docs

- Publishing model: [Publishing overview](/docs/publishing)
- Content system internals: [Content system](/docs/content-system)
- Security & privacy: [Security & privacy](/docs/security-and-privacy)


# Publishing overview
URL: https://app.crispdigital.io/docs/publishing
Section: Publishing

How CRISP Content Engine schedules and publishes content — and where each platform currently sits.

---

CRISP’s publishing model is designed for reliability: **humans approve**, and **workers publish**. Where a platform supports it, CRISP uses native APIs; otherwise CRISP provides an export workflow.

## Who this page is for

- Users scheduling content and expecting it to publish reliably
- Operators configuring cron and troubleshooting publishing failures
- Reviewers validating how CRISP handles tokens, jobs, and data boundaries

## The publishing model

- **Approval creates a publish job** (for native platforms).
- **Workers publish due jobs via cron** (no long-running background jobs in Vercel).
- **Publishing uses the job payload as the source of truth**, not re-reading content stores at publish time.

## Platform support

- **LinkedIn**: native publishing (cron-based)
- **Meta (Facebook + Instagram)**: native publishing (Growth tier and above)
- **𝕏**: export-only (copy/paste)
- **Blog**: export-only (copy/paste)

## Constraints and edge cases

- **Cron is required** for native publishing.
- **Retries are expected**: network and platform API failures happen — jobs move through retry and failed states.
- **Platform limits apply**: daily caps, media requirements, and permission scopes vary by platform.

## Related docs

- LinkedIn publishing: [LinkedIn publishing](/docs/publishing/linkedin)
- Meta publishing: [Meta publishing](/docs/publishing/meta)
- 𝕏 export workflow: [𝕏 publishing](/docs/publishing/x)
- Blog export workflow: [Blog publishing](/docs/publishing/blog)
- Content system (queue guards, retries): [Content system](/docs/content-system)


# LinkedIn publishing
URL: https://app.crispdigital.io/docs/publishing/linkedin
Section: Publishing

Native LinkedIn publishing is cron-based and runs after human approval.

---

LinkedIn is a native publishing platform in CRISP Content Engine: approved content is queued as a publish job and sent to LinkedIn by a scheduled worker.

## Who this page is for

- Users approving/scheduling LinkedIn posts
- Operators setting up cron and troubleshooting status drift

## How it works (at a glance)

- Users connect LinkedIn in **Connections**.
- Approving content creates/updates publish state for LinkedIn.
- A cron worker runs periodically and publishes due items.

## Constraints and edge cases

- **Cron must run** (Vercel does not run background cron).
- **Token expiry happens**: re-authentication may be required if publishing fails with auth errors.
- **Rate limits** and API availability are outside CRISP’s control.

## Troubleshooting checklist

- Confirm cron is calling the LinkedIn due endpoint on schedule.
- Confirm the LinkedIn connection is still valid for the user/workspace.
- If items are stuck, check the app’s publishing status UI and retry workflows.

## Related docs

- Publishing overview: [Publishing overview](/docs/publishing)
- Content system status model: [Content system](/docs/content-system)
- Integrations overview: [Integrations](/docs/integrations)


# Meta publishing (Facebook + Instagram)
URL: https://app.crispdigital.io/docs/publishing/meta
Section: Publishing

Native publishing to Facebook Pages and Instagram Business accounts — available on Growth tier and above.

---

Meta publishing (Facebook Pages + Instagram Business accounts) is **live** in CRISP Content Engine. It is available on the **Growth** plan and above. The integration uses encrypted token storage, cron-based workers, and a Meta-compliant data deletion callback.

## Who this page is for

- Users on Growth (or higher) connecting Facebook and Instagram for autopublish
- Operators configuring cron and troubleshooting Meta publishing jobs
- Support/admins verifying permissions and data handling

## What’s included

- OAuth connection flow (Facebook user → Pages → Instagram Business discovery)
- Selection of **one** default Facebook Page and **one** default Instagram account per workspace
- Cron worker publishes due jobs to Facebook/Instagram
- Data deletion callback endpoint for Meta compliance

## Source of truth and scheduling

- **Publish jobs are immutable** once queued.
- Publishing uses **`publish_jobs.payload_json`** as the source of truth.
- The worker is triggered by cron (`/api/publish/meta-due`) and publishes due jobs at execution time.

## Permissions (Meta)

CRISP requests the permissions required for publishing:

- `pages_show_list`
- `pages_read_engagement`
- `pages_manage_posts`
- `instagram_basic`
- `instagram_content_publish`

## Instagram requirements

- **Instagram publishing requires an image** (image + caption only in the current implementation).
- Reels, Stories, and native scheduling are not included.

## Token storage and encryption

- Meta access tokens are **encrypted at rest** using `META_TOKEN_ENCRYPTION_KEY`.
- Tokens and Meta IDs are **not** written to Airtable.
- Publishing uses Supabase tables for encrypted tokens and destination selections.

## Data deletion callback

CRISP implements a Meta-compliant data deletion endpoint:

- `POST /api/meta/data-deletion`
- Deletes Meta connection data for the user and returns a `url` + `confirmation_code` as required by Meta.

## Constraints and edge cases

- **Business verification** may be required for some Meta apps.
- Phase 1 supports **one Page + one Instagram account per workspace**.
- Instagram publishing failures often relate to missing media, permissions, or an invalid/expired token.

## Related docs

- Publishing overview: [Publishing overview](/docs/publishing)
- Security & privacy: [Security & privacy](/docs/security-and-privacy)
- Integrations: [Integrations](/docs/integrations)


# 𝕏 publishing (export-only)
URL: https://app.crispdigital.io/docs/publishing/x
Section: Publishing

𝕏 posts are currently export-only — CRISP provides copy tools, and you publish manually.

---

𝕏 publishing is **export-only** today. CRISP generates 𝕏-ready copy, validates constraints (like character limits), and provides fast copy tools — but it does not publish to 𝕏’s API yet.

## Who this page is for

- Users working with 𝕏 posts in the approval queue
- Teams standardizing a manual publish workflow for 𝕏

## How the export workflow works

- Generate 𝕏 content as part of your multi-channel batch
- Review and edit in the approval queue
- Use **one-click copy** to copy the post text
- Paste into 𝕏 and publish manually

## Constraints and edge cases

- 𝕏 singles must fit the platform’s character limits.
- Threads may be generated but are export-only.
- Scheduling is disabled for X content in Phase 1 of multi-channel.


Native 𝕏 publishing will be considered once there is sufficient user demand and a stable API path.


## Related docs

- Getting started: [Getting started](/docs/getting-started)
- Content system constraints: [Content system](/docs/content-system)


# Blog publishing (export-only)
URL: https://app.crispdigital.io/docs/publishing/blog
Section: Publishing

Blog articles are currently export-only — CRISP generates and formats, you publish in your CMS.

---

Blog publishing is **export-only** today. CRISP generates draft articles and provides copy tools so you can publish in your CMS (Webflow, WordPress, Ghost, etc.).

## Who this page is for

- Users working with blog drafts in the approval queue
- Operators building a lightweight “copy → paste → publish” blog workflow

## How the export workflow works

- Generate blog content as part of your multi-channel batch
- Review and edit in the approval queue
- Use **one-click copy** to copy the article content
- Paste into your CMS and publish manually

## Constraints and edge cases

- Blog content is not scheduled or auto-published in Phase 1.
- Formatting may need light adjustment depending on your CMS editor.

## Related docs

- Publishing overview: [Publishing overview](/docs/publishing)
- Integrations: [Integrations](/docs/integrations)


# Content system
URL: https://app.crispdigital.io/docs/content-system
Section: Content System

The CRISP content system — intent-driven prompting, approval workflow, scheduling guards, retries, and failure states.

---

This page explains how CRISP is designed to produce reliable outcomes: clear intent, a human approval gate, and deterministic publishing behavior.

## Who this page is for

- Operators who need to understand “why it works this way”
- Support/admins troubleshooting edge cases
- Reviewers validating that publishing is safe and deterministic

## Intent-driven prompting (philosophy)

CRISP content generation is built around **intent** and **constraints**:

- We specify the “why” (goal, audience, offer)
- We preserve brand voice rules and guardrails
- We generate channel-specific drafts (LinkedIn vs 𝕏 vs Meta vs Blog)

## Approval workflow

- Drafts arrive in the approval queue.
- A human can edit and approve.
- Approval is the gate that allows a platform publish flow to start (where supported).

## Scheduling logic and queue guards

CRISP uses queue guards to reduce platform risk and user surprise:

- Minimum spacing rules can be enforced per destination (example: Meta Phase 1 uses per-destination spacing when creating jobs).
- Some platforms are export-only and do not support scheduling yet (𝕏 threads, Blog).

## Status model (publish jobs)

Native publishing uses job states that make failures explicit:

- queued → publishing → published
- queued/retrying → failed (after max attempts)

## Retries and failure states

- Retries use backoff to avoid hammering platform APIs.
- Permanent failures are surfaced clearly so a human can take action.

## Related docs

- Publishing overview: [Publishing overview](/docs/publishing)
- Security & privacy: [Security & privacy](/docs/security-and-privacy)


# Security & privacy
URL: https://app.crispdigital.io/docs/security-and-privacy
Section: Security & Privacy

How CRISP handles tokens, data boundaries, encryption, retention, and deletion.

---

CRISP Content Engine is designed to be safe-by-default: minimize sensitive data exposure, encrypt tokens, and keep publishing deterministic.

## Who this page is for

- Security-minded customers evaluating CRISP
- Operators configuring integrations and environment variables
- Platform reviewers (Meta / LinkedIn) validating data handling patterns

## Data architecture (high level)

CRISP uses multiple systems with clear responsibilities:

- **Airtable**: content records and editorial fields (human-visible)
- **Supabase (Postgres)**: connections, encrypted tokens, publishing jobs, job state
- **Make.com**: content generation workflow engine (webhook-driven)
- **Cloudinary**: image hosting and transformation

## Token storage and encryption

- Platform access tokens are stored server-side.
- Tokens are **encrypted at rest** when stored (example: Meta uses `META_TOKEN_ENCRYPTION_KEY`).
- Tokens and platform IDs are **not written to Airtable** (especially Meta).


When publishing, CRISP uses the stored job payload (e.g. `publish_jobs.payload_json`) rather than re-reading Airtable at publish time.


## Airtable vs Supabase responsibilities

- Airtable supports the editorial workflow (drafts, approval, metadata).
- Supabase is the operational system for:
  - Connection state
  - Destination selection
  - Publish job queue + retries + immutable publish payloads

## Data retention and deletion

- Disconnecting a platform removes connection records and invalidates future jobs.
- Meta requires a dedicated **data deletion callback** endpoint; CRISP implements this as part of the Meta integration.

## Subprocessors and third-party services

Common third parties used by CRISP include:

- Supabase (auth + database)
- Airtable (content tables)
- Make.com (automation/workflows)
- Cloudinary (image hosting)
- Vercel (hosting)
- Stripe (billing)
- Resend (transactional email)
- Platform APIs (LinkedIn, Meta)

## Related docs

- Integrations: [Integrations](/docs/integrations)
- Meta publishing: [Meta publishing](/docs/publishing/meta)


# Integrations
URL: https://app.crispdigital.io/docs/integrations
Section: Integrations

How CRISP integrates with Airtable, Make.com, Cloudinary, and platform APIs.

---

CRISP Content Engine integrates with a small number of best-in-class services to keep the product fast, reliable, and secure.

## Who this page is for

- Operators setting up CRISP for a workspace
- Developers/support debugging data flow across systems

## Integration model (high level)

- **Airtable**: editorial content records
- **Make.com**: workflow orchestration for generation and callbacks
- **Cloudinary**: image hosting (and required media URLs for platforms)
- **Platform APIs**: publishing connections and publishing workers

## Constraints and edge cases

- Secrets (API keys, cron secrets, token encryption keys) must live in environment variables.
- Webhook payloads must stay stable for Make.com scenarios to remain reliable.

## Related docs

- Airtable: [Airtable integration](/docs/integrations/airtable)
- Cloudinary: [Cloudinary and images](/docs/integrations/cloudinary)
- Make.com: [Make.com and webhooks](/docs/integrations/make)
- Security & privacy: [Security & privacy](/docs/security-and-privacy)


# Airtable integration
URL: https://app.crispdigital.io/docs/integrations/airtable
Section: Integrations

Airtable powers the editorial content workflow in CRISP Content Engine.

---

Airtable is used as the human-friendly system for content records and editorial fields.

## Who this page is for

- Operators configuring Airtable bases/tables/fields
- Support and admins troubleshooting missing or malformed queue items

## What Airtable is used for

- Content records (drafts, approval status, copy, scheduling fields)
- Brand and strategy references
- Operational visibility for humans

## What Airtable is NOT used for

- **Sensitive tokens** (e.g. Meta access tokens)
- Publishing source-of-truth payloads at publish time (native jobs publish from stored job payloads)


This boundary is deliberate: Airtable can change after approval, and publishing must remain deterministic.


## Constraints and edge cases

- Field name and schema changes can break ingestion unless updated in CRISP.
- Airtable rate limits apply; CRISP avoids unnecessary reads during publish flows.

## Related docs

- Integrations overview: [Integrations](/docs/integrations)
- Security & privacy: [Security & privacy](/docs/security-and-privacy)


# Cloudinary and images
URL: https://app.crispdigital.io/docs/integrations/cloudinary
Section: Integrations

Image hosting requirements for CRISP, including Instagram publishing constraints.

---

CRISP uses Cloudinary to host images and provide stable, HTTPS-accessible URLs for platforms that require media.

## Who this page is for

- Users uploading images for posts
- Operators validating that image URLs are stable and platform-compatible

## Why Cloudinary is used

- Stable, cache-friendly media URLs
- HTTPS by default
- Suitable for platform API requirements (e.g. Instagram image publishing)

## Constraints and edge cases

- Instagram publishing requires an image (Phase 1 Meta publishing).
- Upload size and format constraints may apply (JPG/PNG/WebP).
- The exact image requirements can vary by platform and can change over time.

## Related docs

- Meta publishing: [Meta publishing](/docs/publishing/meta)
- Integrations overview: [Integrations](/docs/integrations)


# Make.com and webhooks
URL: https://app.crispdigital.io/docs/integrations/make
Section: Integrations

How CRISP uses Make.com for generation workflows, callbacks, and idempotency.

---

CRISP uses Make.com to orchestrate generation workflows. The app triggers Make scenarios via webhooks, and Make calls back into the app when work is complete.

## Who this page is for

- Operators building or updating Make scenarios
- Developers/support troubleshooting generation and callback issues

## How the flow works

- CRISP triggers generation by sending a payload to Make.
- Make generates content and writes results to Airtable.
- Make calls back to CRISP endpoints to confirm completion and update progress.

## Key constraints

- Webhook payloads must remain stable for scenarios to keep working.
- Callbacks must be idempotent to avoid double-counting usage.
- Environment variables must be set correctly (webhook URLs, secrets).

## Common failure scenarios

- Wrong webhook URL or missing secret header
- Airtable schema drift (field/table renamed)
- Make scenario disabled or failing mid-run

## Related docs

- Getting started: [Getting started](/docs/getting-started)
- Airtable integration: [Airtable integration](/docs/integrations/airtable)
- Content system: [Content system](/docs/content-system)


# FAQ
URL: https://app.crispdigital.io/docs/faq
Section: FAQ

Fast answers to common CRISP Content Engine questions and failure scenarios.

---

This page covers the most common questions we see during onboarding and platform setup.

## Who this page is for

- New users learning platform constraints
- Operators troubleshooting publishing and generation
- Reviewers validating Meta requirements and Phase 1 constraints

## Meta publishing (Growth and above)

- Meta (Facebook + Instagram) publishing is **live** and available on the Growth plan and above.
- Business verification may be required depending on your Meta app and permissions.
- CRISP provides a compliant data deletion callback endpoint for Meta App Review.

## Instagram publishing requirements

- Phase 1 Instagram publishing is **image + caption**.
- If there is no image attached, the job cannot publish to Instagram.

## Why 𝕏 is manual (for now)

- 𝕏 publishing is currently export-only to keep the product stable while usage grows.
- CRISP provides copy tools and validation, but does not publish via API yet.

## Common failure scenarios (and what to check)

### “Nothing is publishing”

- Confirm your cron provider is calling the due endpoints on schedule.
- Confirm the cron secret header matches (`X-Cron-Secret`).
- Confirm publishing is enabled for that platform (Meta requires Growth plan or higher and a connected Page/Instagram).

### “Meta connected but posts fail”

- Confirm the selected Page/Instagram account is set.
- Confirm tokens are valid and permissions were granted.
- Confirm the post meets platform requirements (Instagram needs an image).

### “Generated content isn’t showing in the queue”

- Confirm the Make scenario ran successfully.
- Confirm Airtable schema hasn’t changed (field names/tables).
- Confirm callbacks are reachable from Make (no auth required, correct URL).

## Related docs

- Publishing overview: [Publishing overview](/docs/publishing)
- Meta publishing: [Meta publishing](/docs/publishing/meta)
- Integrations: [Integrations](/docs/integrations)


# Changelog
URL: https://app.crispdigital.io/docs/changelog
Section: Changelog

Release notes for CRISP Content Engine (coming soon).

---

This changelog will track product changes over time in a predictable, reviewable format.

## Who this page is for

- Customers who want to understand what changed and why
- Operators validating rollout timing for new publishing features

## Format (future)

- Date
- Summary
- Added / Changed / Fixed
- Notes for operators (migrations, env vars, cron changes)


Changelog entries will be added as features ship. No entries yet.