Notion Database Sync
Create Notion pages from app records with properties mapped and long content split safely.
What it adds
A sync that writes app records as pages in a selected database, with mapped properties and structured body content.
What your agent is told to do
5
What your agent is told to do
5-
1
Confirm at setup, and again before each run, that the selected database is still shared with the integration, because sharing is granted per page and can be withdrawn without any notice to the app.
-
2
Read the database's property identifiers and types from the destination and map each app field by identifier, not by the label a user is free to rename.
-
3
Format each property according to its own type, and treat select, status, relation, and people properties as closed sets. Resolve the value before the page write; do not send an unknown value hoping the provider will create it.
-
4
Split body content into blocks that respect the provider's per-block and per-request limits, and preserve order across the several requests a long record will take.
-
5
Carry a stable app identifier in a dedicated property and look it up before writing, so a retried job updates the existing page rather than creating a second one. This is the same idempotency contract Airtable Record Sync uses; implement it once and share it.
Edge cases it handles
8
Edge cases it handles
8- Access is granted by sharing a database with the integration, and any workspace member can unshare it. The sync must detect the loss and pause with a message rather than retrying against a database it can no longer see.
- Each property type accepts a different shape. Sending plain text to a date, number, or multi-select property fails, and the error often names only the property, so validate types before the request.
- Relation, people, select, and status properties reject values that do not already exist. Decide per field whether an unknown value is created, mapped to a fallback, or reported, and make that a setting rather than a guess in code.
- Long records exceed the block and request limits. Split content into valid blocks, send them in ordered batches, and make a partially written page recoverable rather than leaving half a record behind.
- Retries and redelivered events create a second page unless the sync looks up its own key first. Query by the identifier property before every create.
- Properties get renamed and deleted by whoever owns the database. A mapping stored by name breaks invisibly, while a mapping stored by identifier survives a rename and fails loudly on a delete.
- Rate limits are enforced per integration, so a backfill competes with ordinary writes. Pace the backfill and let user-triggered syncs take priority.
- Content copied out of the app may carry markup the destination interprets. Convert it into the destination's block structure rather than pasting raw markup into a text property.
Definition of done
8
Definition of done
8- The sync verifies the database is still shared before each run and pauses with a clear message when it is not.
- Properties are mapped by identifier, so renaming a property in the destination does not break the sync.
- Each property type receives correctly formatted values, and unknown option values follow a configured rule.
- Long content is split into valid blocks and written in order, with partial writes recoverable.
- A stable key property makes repeated runs idempotent, and no retry creates a duplicate page.
- Backfills respect the provider's rate limits without blocking user-triggered syncs.
- The feature matches the existing design system.
- No existing functionality is broken.
Related features
Contentful Entry Sync
Contentful Entry Sync
Create and update Contentful entries against the connected space, environment, and locale.
What it does
A sync that writes app records into Contentful entries using the connected space's content model and locale configuration.
How it works
- 1 Have the operator select the space, the environment, and the default locale during connection, and store all three, because writing production content into a sandbox environment is an easy and expensive mistake.
- 2 Fetch the content model at connection time and again before each sync run, map app fields to entry fields by their stable identifiers, and validate the value type against the model before attempting the write.
- 3 Write every localized field under an explicit locale rather than relying on whatever the space treats as default, and state clearly which locales the app owns and which it leaves to editors.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/contentful-entry-sync
Typesense Index Sync
Typesense Index Sync
Keep a search index in step with your records so results stay current and correctly scoped.
What it does
A one-way sync from the app's records into a Typesense collection, with server-enforced tenant filtering and a rebuild path.
How it works
- 1 Identify every model users search today and declare the indexed shape explicitly: the fields people type into, the fields they filter and sort on, and nothing more. Treat a missing or newly added field as absent rather than letting the document fail validation.
- 2 Reuse the app's existing background-job system to push creates, updates, and deletes. Do not index inside the request cycle; a slow or unreachable search service must never make a save fail or hang.
- 3 Build the tenant, workspace, and permission filter on the server at query time and attach it to every search. Do not accept a filter, collection name, or scope value from a client-supplied parameter, because anything the browser can set the browser can change.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/typesense-index-sync
Mapbox Geocoding
Mapbox Geocoding
Convert typed locations into coordinates and structured addresses the app can use.
What it does
Forward and reverse geocoding with confirmed matches, controlled result bias, and coordinate precision rules.
How it works
- 1 Separate the two directions explicitly: text to coordinates for entry and search, coordinates to address for a dropped pin or a device location. They take different inputs, carry different confidence, and fail differently.
- 2 Present ambiguous matches as a list the user confirms. Silently taking the first result is how a record ends up in the wrong country with nobody noticing for months.
- 3 Bias results toward a location only when the app genuinely knows one, such as a workspace address or the map viewport the user is already looking at.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/mapbox-geocoding
How it works
-
1
Copy the link
Grab the Markdown instruction URL for this feature.
-
2
Give it to your AI
Paste it into Claude Code, Cursor, v0, Lovable — whatever you build with.
-
3
It inspects, then implements
Your agent reads your existing app first, then adds the feature to fit it.
Works with your stack
These instructions are written to adapt. They tell the agent to detect your framework, match your existing design system, and reuse what you already have — rather than assuming a particular stack.
Need it tighter than that? Customize the feature and tell it exactly what you're running.