agentleFS
JustineDevsJustineDevs/premortem/.agents/skills/building-nango-functions-locally/SKILL.md

building-nango-functions-locally

Skill2 starsChanged 5 months agoApache-2.0View on GitHub

Builds Nango Functions in a checked-out Zero YAML TypeScript Nango project using local files, index.ts registration, nango dryrun, generated tests, and optional nango deploy via CLI. Use when creating, updating, validating, testing, or deploying Nango actions or syncs locally in a repo. This content overlaps with building-nango-functions but adds CLI workflow details, so load this instead of building-nango-functions whenever a local project, local files, Nango root, CLI, dryrun, generated tests, or nango deploy workflow is indicated.

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

49 rules. About 0 are common, 17 are distinctive, and 32 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈17

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 32

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the active workflow skill for compile, dryrun, test, and deploy mechanics

    line 13
  2. Very specific:

    If the task is a sync, read references/syncs.md before writing code and state one of these paths first:

    line 17
  3. Very specific:

    Read references/actions.md before writing code

    line 41
  4. Very specific:

    Test input JSON if the active workflow will validate or dryrun the action (use {} for no-input actions)

    line 61
  5. Distinctive:

    Delete strategy (deleted-record endpoint/webhook, or why full refresh is required)

    line 68
  6. Distinctive:

    If proposing a full refresh, the exact provider limitation that blocks checkpoints from the docs/sample response

    line 69
  7. Distinctive:

    Use the Nango HTTP API for connection lookup, credentials, and proxy calls outside function code.

    line 80
  8. Distinctive:

    Add an API doc link comment above each provider call.

    line 81
  9. Very specific:

    Do not set deprecated function definition routing fields: action endpoint and sync endpoints.

    line 85
  10. Very specific:

    Default to nango.paginate(...) + nango.batchSave(...).

    line 93
  11. Very specific:

    Prefer batchDelete() when the provider returns deletions, tombstones, or delete webhooks.

    line 94
  12. Distinctive:

    Use full refresh only if the provider cannot return changes, deletions, or resume state, or if the dataset is tiny.

    line 95
  13. Very specific:

    Never combine trackDeletesStart() / trackDeletesEnd() with changed-only checkpoints (modified_after, updated_after, changed-records …

    line 98
  14. Very specific:

    If a sync requires metadata (e.g.

    line 100
  15. Distinctive:

    Prefer explicit field names.

    line 105
  16. Very specific:

    Add .describe() examples for IDs, timestamps, enums, and URLs.

    line 106
  17. Very specific:

    Avoid any; use inline mapping types.

    line 107
  18. Very specific:

    Use nango.zodValidateInput() only when you need custom validation or logging.

    line 109
  19. Very specific:

    Default non-required inputs to .optional().

    line 113
  20. Very specific:

    Use .nullable() only when null has meaning, usually clear-on-update.

    line 114
  21. Very specific:

    Default generated schemas to .optional() for non-required inputs and normalized outputs.

    line 117
  22. Very specific:

    Prefer .nullable() over z.union([z.null(), T]) or z.union([T, z.null()]).

    line 118
  23. Very specific:

    Use explicit suffixes in the API's majority casing: IDs (user_id, userId), names (channel_name, channelName), emails (user_email, …

    line 124
  24. Very specific:

    If the API is snake_case, use user_id instead.

    line 142
  25. Distinctive:

    If web fetching returns incomplete docs (JS-rendered):

    line 160
  26. Distinctive:

    Ask the user for a sample response

    line 161
  27. Distinctive:

    Use existing Nango actions or syncs in the workspace as a pattern when they exist

    line 162
  28. Distinctive:

    Use the skill-specific validation or dryrun workflow until it passes

    line 163
  29. Very specific:

    Read the matching reference file: references/actions.md or references/syncs.md.

    line 167
  30. Very specific:

    Create or update the function under {integrationId}/actions/ or {integrationId}/syncs/, apply the shared schema and casing rules, then …

    line 171
  31. Distinctive:

    If validation cannot pass, stop and report the missing external state or inputs.

    line 173
  32. Very specific:

    After validation passes, run nango dryrun ...

    line 174
  33. Very specific:

    If you see YAML PROJECT DETECTED:

    line 187
  34. Distinctive:

    Do not attempt to mix YAML and TypeScript.

    line 190
  35. Distinctive:

    Do not create files until you confirm the Nango root:

    line 196
  36. Distinctive:

    Do not use absolute paths as a workaround

    line 205
  37. Distinctive:

    All file paths must be relative to the Nango root.

    line 207
  38. Very specific:

    All actions and syncs must be imported in index.ts

    line 229
  39. Very specific:

    Use side-effect imports only.

    line 233
  40. Very specific:

    Run nango dryrun ... --validate -e dev --no-interactive --auto-confirm until it passes.

    line 246
  41. Distinctive:

    If validation cannot pass, stop and state the missing external state or inputs required.

    line 249
  42. Very specific:

    After validation passes, run nango dryrun ...

    line 250
  43. Very specific:

    Run nango generate:tests, then npm test.

    line 251
  44. Very specific:

    Treat <script-name>.test.json as generated output.

    line 279
  45. Very specific:

    If mocks are wrong or stale, fix the code and re-record with --save.

    line 280
  46. Very specific:

    Do not hard-code error payloads in *.test.json.

    line 281
  47. Very specific:

    Use --integration-id <integration-id> when script names overlap across integrations.

    line 283
  48. Very specific:

    Prefer --checkpoint for new incremental syncs.

    line 284
  49. Very specific:

    If nango is not on PATH, use npx nango ....

    line 285

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Build Nango Functions Locally

Build deployable Nango actions and syncs in a checked-out Nango project with the local CLI validation and test workflow.

Implementation Scope

  • Build or modify a Nango function implementation
  • Build an action in Nango with createAction()
  • Build a sync in Nango with createSync()
  • Use the active workflow skill for compile, dryrun, test, and deploy mechanics

Sync Strategy Gate (required before writing code)

If the task is a sync, read references/syncs.md before writing code and state one of these paths first:

  • Checkpoint plan:
    • change source (updated_at, modified_since, changed-records endpoint, cursor, page token, offset/page, since_id, or webhook)
    • checkpoint schema
    • how the checkpoint changes the provider request or resume state
    • whether the request still walks the full dataset or returns changed rows only
    • delete strategy
  • Full refresh blocker:
    • exact provider limitation from the docs or sample payloads
    • why checkpoints cannot work here

Invalid sync implementations:

  • full refresh because it is simpler
  • saveCheckpoint() without getCheckpoint()
  • reading or saving a checkpoint without using it in request params or pagination state
  • using syncType: 'incremental' or nango.lastSyncDate in a new sync
  • using trackDeletesStart() / trackDeletesEnd() with a changed-only checkpoint (modified_after, updated_after, changed-records endpoint). Those requests omit unchanged rows, so trackDeletesEnd() will falsely delete them.
  • using trackDeletesStart() / trackDeletesEnd() in an incremental sync that already has explicit deleted-record events

Choose the Path

Action:

  • One-time request, user-triggered, built with createAction()
  • Read references/actions.md before writing code

Sync:

  • Scheduled or webhook-driven cache updates built with createSync()
  • Complete the Sync Strategy Gate first
  • Read references/syncs.md before writing code

Required Inputs (Ask User if Missing)

Always:

  • Integration ID (provider name)
  • Script/function name (kebab-case)
  • API reference URL or sample response
  • Connection ID if the active workflow will validate or dryrun the function

Action-specific:

  • Use case summary
  • Input parameters
  • Output fields
  • Metadata JSON if required
  • Test input JSON if the active workflow will validate or dryrun the action (use {} for no-input actions)

Sync-specific:

  • Model name (singular, PascalCase)
  • Frequency (every hour, every 5 minutes, etc.)
  • Checkpoint schema (timestamp, cursor, page token, offset/page, since_id, or composite)
  • How the checkpoint changes the provider request or resume state
  • Delete strategy (deleted-record endpoint/webhook, or why full refresh is required)
  • If proposing a full refresh, the exact provider limitation that blocks checkpoints from the docs/sample response
  • Metadata JSON if required (team_id, workspace_id)

If any required external values are missing, ask a targeted question after checking the repo and provider docs. For syncs, choose a checkpoint plus deletion strategy whenever the provider supports one. If you cannot find a viable checkpoint strategy, state exactly why before writing a full refresh.

Non-Negotiable Rules

Shared platform constraints

  • Nango functions use createAction() / createSync().
  • You cannot add arbitrary packages. Use relative imports only when the chosen workflow supports them; built-ins include zod, crypto/node:crypto, and url/node:url.
  • Use the Nango HTTP API for connection lookup, credentials, and proxy calls outside function code. Do not invent CLI token or connection commands.
  • Add an API doc link comment above each provider call.
  • Action outputs cannot exceed 2MB.
  • File uploads and downloads cannot be implemented as actions (sandboxed runtime: no fs, no axios, 2 MB output limit). Use a proxy script in {integration}/proxy/ with @nangohq/node instead — see references/actions.md.
  • HTTP retries default to 0; set retries deliberately. Treat 3 as the normal maximum; for sync provider calls, values above 3 are effectively forbidden unless docs prove they are safe and necessary. Avoid retries for non-idempotent writes unless the API supports idempotency.
  • Do not set deprecated function definition routing fields: action endpoint and sync endpoints. Trigger actions by action name through the SDK/API, and consume sync records through the records API.

Sync rules

  • Sync records need a stable string id.
  • New syncs should define a checkpoint schema, call nango.getCheckpoint() first, and nango.saveCheckpoint() after each page or batch.
  • A checkpoint is valid only if it changes the request or resume state (since, updated_after, cursor, page_token, offset, page, since_id, etc.). Saving one without using it is not incremental sync.
  • New syncs must not use syncType: 'incremental' or nango.lastSyncDate.
  • Default to nango.paginate(...) + nango.batchSave(...). Avoid manual while (true) loops when cursor, link, or offset pagination fits.
  • Prefer batchDelete() when the provider returns deletions, tombstones, or delete webhooks.
  • Use full refresh only if the provider cannot return changes, deletions, or resume state, or if the dataset is tiny.
  • For full refresh, cite the exact provider limitation from docs or payloads. "It is easier" is not enough.
  • deleteRecordsFromPreviousExecutions() is deprecated. For full refresh, call trackDeletesStart() before fetch/save and trackDeletesEnd() only after a successful full fetch/save.
  • Never combine trackDeletesStart() / trackDeletesEnd() with changed-only checkpoints (modified_after, updated_after, changed-records endpoints, etc.). They omit unchanged rows, so trackDeletesEnd() would delete them.
  • Checkpointed full refreshes are still full refreshes. Call trackDeletesEnd() only in the run that finishes the full window.
  • If a sync requires metadata (e.g. team_id, workspace_id, guild_id), set autoStart: false. The sync cannot run until the caller has set the metadata, so starting it automatically would fail.

Conventions

  • Match field casing to the external API. Passthrough fields keep provider casing; non-passthrough fields should use the majority casing of that API.
  • Prefer explicit field names.
  • Add .describe() examples for IDs, timestamps, enums, and URLs.
  • Avoid any; use inline mapping types.
  • List actions should expose cursor plus a next-cursor field in the majority casing of that API (next_cursor, nextCursor, etc.).
  • Use nango.zodValidateInput() only when you need custom validation or logging; otherwise rely on schemas plus the chosen validation workflow.

Schema Semantics

  • Default non-required inputs to .optional().
  • Use .nullable() only when null has meaning, usually clear-on-update; add .optional() when callers may omit the field too.
  • Raw provider schemas should match the provider: .optional() for omitted fields, .nullable() for explicit null, .nullish() only when the provider truly does both.
  • Final action outputs and normalized sync models should prefer .optional() and normalize upstream null to omission unless null matters.
  • Default generated schemas to .optional() for non-required inputs and normalized outputs; widen only when the upstream contract justifies it.
  • Prefer .nullable() over z.union([z.null(), T]) or z.union([T, z.null()]).
  • Return null only when the output schema allows it.
  • z.object() strips unknown keys by default. For provider pass-through use z.object({}).passthrough(), z.record(z.unknown()), or z.unknown() with minimal refinements.

Field Naming and Casing Rules

  • Use explicit suffixes in the API's majority casing: IDs (user_id, userId), names (channel_name, channelName), emails (user_email, userEmail), URLs (callback_url, callbackUrl), and timestamps (created_at, createdAt).

Mapping example (API expects a different parameter name):

const InputSchema = z.object({
    userId: z.string()
});

const config: ProxyConfiguration = {
    endpoint: 'users.info',
    params: {
        user: input.userId
    },
    retries: 3
};

If the API is snake_case, use user_id instead. The goal is API consistency.

References

  • Action patterns, CRUD examples, metadata usage, and ActionError examples: references/actions.md
  • Sync patterns, concrete checkpoint examples, delete strategies, and full refresh fallback: references/syncs.md

Useful Nango docs (quick links)

When API Docs Do Not Render

If web fetching returns incomplete docs (JS-rendered):

  • Ask the user for a sample response
  • Use existing Nango actions or syncs in the workspace as a pattern when they exist
  • Use the skill-specific validation or dryrun workflow until it passes

Workflow (required)

  1. Decide whether this is an action or a sync.
  2. Read the matching reference file: references/actions.md or references/syncs.md.
  3. For syncs, inspect provider docs or payloads for checkpoints and deletes, decide whether the endpoint returns full data or changed rows, and complete the Sync Strategy Gate.
  4. Gather required inputs and external values. For connection lookup, credentials, or discovery, use the Nango HTTP API.
  5. Confirm this is a Zero YAML TypeScript project (no nango.yaml) and that you are in the Nango root (.nango/ exists).
  6. Create or update the function under {integrationId}/actions/ or {integrationId}/syncs/, apply the shared schema and casing rules, then register it in index.ts.
  7. Validate with nango dryrun ... --validate -e dev --no-interactive --auto-confirm.
  8. If validation cannot pass, stop and report the missing external state or inputs.
  9. After validation passes, run nango dryrun ... --save, then nango generate:tests, then npm test.
  10. Deploy with nango deploy dev only when requested.

Preconditions (Do Before Writing Code)

Confirm TypeScript Project (No nango.yaml)

This skill only supports TypeScript projects using createAction() / createSync().

ls nango.yaml 2>/dev/null && echo "YAML PROJECT DETECTED" || echo "OK - No nango.yaml"

If you see YAML PROJECT DETECTED:

  • Stop immediately.
  • Tell the user to upgrade to the TypeScript format first.
  • Do not attempt to mix YAML and TypeScript.

Reference: https://nango.dev/docs/implementation-guides/platform/migrations/migrate-to-zero-yaml

Verify Nango Project Root

Do not create files until you confirm the Nango root:

ls -la .nango/ 2>/dev/null && pwd && echo "IN NANGO PROJECT ROOT" || echo "NOT in Nango root"

If you see NOT in Nango root:

  • cd into the directory that contains .nango/
  • Re-run the check
  • Do not use absolute paths as a workaround

All file paths must be relative to the Nango root. Creating files with extra prefixes while already in the Nango root will create nested directories that break the build.

Project Structure and Naming

./
|-- .nango/
|-- index.ts
|-- hubspot/
|   |-- actions/
|   |   `-- create-contact.ts
|   `-- syncs/
|       `-- fetch-contacts.ts
`-- slack/
    `-- actions/
        `-- post-message.ts
  • Provider directories: lowercase (hubspot, slack)
  • Action files: kebab-case (create-contact.ts)
  • Sync files: kebab-case (many teams use a fetch- prefix, but it is optional)
  • One function per file
  • All actions and syncs must be imported in index.ts

Register scripts in index.ts (required)

Use side-effect imports only. Include the .js extension.

// index.ts
import './github/actions/get-top-contributor.js';
import './github/syncs/fetch-issues.js';

Symptom of incorrect registration: the file compiles but you see No entry points found in index.ts... or the function never appears.

Dryrun, Mocks, and Tests (required)

Required loop:

  1. Run nango dryrun ... --validate -e dev --no-interactive --auto-confirm until it passes.
  2. Actions: always pass --input '{...}' (use --input '{}' for no-input actions).
  3. Syncs: use --checkpoint '{...}' when you need to simulate a resumed run.
  4. If validation cannot pass, stop and state the missing external state or inputs required.
  5. After validation passes, run nango dryrun ... --save -e dev --no-interactive --auto-confirm to generate <script-name>.test.json.
  6. Run nango generate:tests, then npm test.

Examples:

# Validate an action
nango dryrun <action-name> <connection-id> --validate -e dev --no-interactive --auto-confirm --input '{"key":"value"}'

# Validate a no-input action
nango dryrun <action-name> <connection-id> --validate -e dev --no-interactive --auto-confirm --input '{}'

# Validate a sync
nango dryrun <sync-name> <connection-id> --validate -e dev --no-interactive --auto-confirm

# Validate a resumed sync with a checkpoint
nango dryrun <sync-name> <connection-id> --validate -e dev --no-interactive --auto-confirm --checkpoint '{"updated_after":"2024-01-15T00:00:00Z"}'

# Record action mocks after validation passes
nango dryrun <action-name> <connection-id> --save -e dev --no-interactive --auto-confirm --input '{"key":"value"}'

# Record sync mocks after validation passes
nango dryrun <sync-name> <connection-id> --save -e dev --no-interactive --auto-confirm

# Stub metadata when needed
nango dryrun <script-name> <connection-id> --validate -e dev --no-interactive --auto-confirm --metadata '{"team_id":"123"}'

Hard rules:

  • Treat <script-name>.test.json as generated output. Never create, edit, rename, or move it.
  • If mocks are wrong or stale, fix the code and re-record with --save.
  • Do not hard-code error payloads in *.test.json; use a Vitest test with vi.spyOn(...) for 404, 401, 429, or timeout cases.
  • Connection ID is the second positional argument; do not use --connection-id.
  • Use --integration-id <integration-id> when script names overlap across integrations.
  • Prefer --checkpoint for new incremental syncs; --lastSyncDate is a legacy pattern.
  • If nango is not on PATH, use npx nango ....
  • CLI upgrade prompts can block automation; set NANGO_CLI_UPGRADE_MODE=ignore if needed.

Reference: https://nango.dev/docs/implementation-guides/platform/functions/testing

Deploy (Optional)

Deploy functions to an environment in your Nango account:

nango deploy dev

# Deploy only one function
nango deploy --action <action-name> dev
nango deploy --sync <sync-name> dev

Reference: https://nango.dev/docs/implementation-guides/use-cases/actions/implement-an-action

Final Checklists

Action:

  • Nango root verified
  • references/actions.md was used for the action pattern
  • Schemas and types are clear, and missing-value rules match the provider versus normalized contract
  • createAction() includes input, output, and scopes when required; deprecated endpoint is omitted
  • Fields use passthrough casing or the API's majority casing
  • Provider call includes an API doc link comment and intentional retries
  • nango.ActionError is used for expected failures
  • Registered in index.ts
  • Dryrun succeeds with --validate -e dev --no-interactive --auto-confirm --input '{...}'
  • <action-name>.test.json was generated by nango dryrun ... --save after --validate
  • nango generate:tests ran and npm test passes

Sync:

  • Nango root verified
  • references/syncs.md was used for the sync pattern
  • Models map is defined, ids are stable strings, and normalized models prefer .optional() unless null matters
  • Incremental was chosen first, with a checkpoint schema unless full refresh is explicitly justified from docs or payloads
  • nango.getCheckpoint() is read at the start and nango.saveCheckpoint() runs after each page or batch
  • Checkpoint data changes the provider request or resume state (since, updated_after, cursor, page_token, offset, page, since_id, etc.)
  • Changed-only checkpoint syncs (modified_after, updated_after, changed-records endpoint) do not use trackDeletesStart() / trackDeletesEnd()
  • If checkpoints were not used, the response explains exactly why no viable checkpoint strategy exists
  • Raw provider schemas model omitted versus null correctly, and fields use passthrough casing or the API's majority casing
  • nango.paginate() is used unless the API truly cannot fit Nango's paginator
  • Provider API calls use retries: 3; no sync retry value exceeds 3 without a documented exception
  • Deletion strategy matches the sync type: batchDelete() for incremental only when the provider returns explicit deletions; otherwise full-refresh fallback uses trackDeletesStart() before fetch/save and trackDeletesEnd() only after a successful full fetch plus save
  • Metadata handled if required
  • Registered in index.ts
  • Dryrun succeeds with --validate -e dev --no-interactive --auto-confirm
  • <sync-name>.test.json was generated by nango dryrun ... --save after --validate
  • nango generate:tests ran and npm test passes

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.