@lhi/n8m 0.2.3 → 0.3.0

package/docs/N8N_NODE_REFERENCE.md

@@ -0,0 +1,369 @@
+# n8n Node Reference for AI Models
+
+This document serves as a static reference for common n8n nodes when the live
+fetch from the n8n instance fails. It covers some of the most frequently used
+nodes and their parameters.
+
+## Core Nodes
+
+### 1. Manual Trigger (n8n-nodes-base.start)
+
+- **Description**: The starting point for manual execution.
+- **Parameters**: (None usually required)
+
+### 2. HTTP Request (n8n-nodes-base.httpRequest)
+
+- **Description**: Send HTTP requests to any API.
+- **Key Parameters**:
+  - `method`: HTTP method (GET, POST, PUT, DELETE, etc.)
+  - `url`: The endpoint URL.
+  - `authentication`: "none", "predefinedCredentialType".
+  - `sendBody`: (boolean) Whether to send a body (required for POST/PUT).
+  - `body`: { `contentType`: "json"|"form-data"|"binary", `content`: { ... } }
+  - `sendHeaders`: (boolean)
+  - `headerParameters`: List of { name, value }.
+
+### 3. Slack (n8n-nodes-base.slack)
+
+- **Description**: Send messages, manage channels, users, files, and reactions in Slack.
+- **Resources & Operations**:
+  - `message`: post | update | delete | get | getAll | search
+  - `channel`: create | archive | unarchive | get | getAll | history | invite | join | kick | leave | open | rename | replies | setPurpose | setTopic
+  - `file`: upload | get | getAll
+  - `reaction`: add | remove | get
+  - `user`: get | getAll | getPresence
+  - `userGroup`: create | disable | enable | get | getAll
+- **Core Parameters** (message.post):
+  - `channel`: Channel name (`#general`) or ID (`C1234567890`). Required.
+  - `text`: Plain-text fallback. Required when no blocks/attachments present. Supports mrkdwn if `otherOptions.mrkdwn=true`.
+  - `blocksUi`: **Block Kit blocks** as a JSON array string. See Block Kit section below.
+  - `attachments`: Legacy Slack attachment array as JSON string. Prefer `blocksUi` for new workflows.
+- **otherOptions** (collection, all optional):
+  - `thread_ts`: Timestamp of parent message — posts as a thread reply. Source from `$json.ts` on a previous Slack response.
+  - `reply_broadcast`: `true` to also broadcast a thread reply to the channel.
+  - `mrkdwn`: `true` to enable Markdown in `text`.
+  - `username`: Override the bot display name for this post.
+  - `icon_emoji`: Override bot icon with an emoji (`:robot_face:`).
+  - `icon_url`: Override bot icon with an image URL.
+  - `unfurl_links` / `unfurl_media`: Control link/media previews.
+
+#### Block Kit (`blocksUi`) — CRITICAL NOTES
+
+`blocksUi` must be a **valid JSON array string** at runtime. n8n parses it with
+`JSON.parse()` — if the expression produces anything invalid the node throws
+`"Parameter 'blocksUi' could not be parsed"`.
+
+**Rules:**
+1. The entire value must be a JSON array: `[{...}, {...}]`
+2. n8n expressions (`={{ ... }}`) go **inside string values**, never at the array wrapper level.
+3. If a referenced field might be `null`/`undefined`, guard it: `={{ $json.body.content ?? '' }}`
+4. Keep block text under 3000 chars. Max 50 blocks per message.
+
+**Supported block types**: `section`, `header`, `divider`, `image`, `actions`, `context`, `input`
+
+**Minimal working examples:**
+
+```json
+// Simple text section
+[{"type":"section","text":{"type":"mrkdwn","text":"={{ $json.body.content }}"}}]
+
+// Header + body
+[
+  {"type":"header","text":{"type":"plain_text","text":"={{ $json.body.title }}","emoji":true}},
+  {"type":"section","text":{"type":"mrkdwn","text":"={{ $json.body.content }}"}},
+  {"type":"divider"}
+]
+
+// Section with two fields side-by-side
+[{"type":"section","fields":[
+  {"type":"mrkdwn","text":"*Status:*\n={{ $json.status }}"},
+  {"type":"mrkdwn","text":"*Priority:*\n={{ $json.priority }}"}
+]}]
+
+// Action buttons
+[{"type":"actions","elements":[
+  {"type":"button","text":{"type":"plain_text","text":"Approve"},"style":"primary","action_id":"approve","value":"={{ $json.id }}"},
+  {"type":"button","text":{"type":"plain_text","text":"Reject"},"style":"danger","action_id":"reject","value":"={{ $json.id }}"}
+]}]
+
+// Context (small grey text)
+[{"type":"context","elements":[{"type":"mrkdwn","text":"Posted by n8m • {{ $now.toISO() }}"}]}]
+```
+
+**Thread reply example:**
+```json
+{
+  "resource": "message",
+  "operation": "post",
+  "channel": "#alerts",
+  "text": "={{ $json.summary }}",
+  "otherOptions": {
+    "thread_ts": "={{ $json.ts }}",
+    "reply_broadcast": false
+  }
+}
+```
+
+#### Trigger (n8n-nodes-base.slackTrigger)
+
+Receives Slack events via webhooks. Key `triggerOn` values:
+- `message` — new message in a channel
+- `appMention` — bot is @mentioned
+- `reactionAdded` / `reactionRemoved`
+- `teamJoin` — new user joins workspace
+- `channelCreated` — new public channel
+
+### 4. Set (n8n-nodes-base.set)
+
+- **Description**: Manipulate data or create new variables.
+- **Key Parameters**:
+  - `values`: Array of { `name`, `type` (string/number/boolean), `value` }.
+  - `options`: { `keepOnlySet`: boolean }.
+
+### 5. IF (n8n-nodes-base.if)
+
+- **Description**: Conditional logic (True/False branches).
+- **Key Parameters**:
+  - `conditions`: { `string`|`number`|`boolean`: [ { `value1`, `operation`,
+    `value2` } ] }
+  - Common operations: `equals`, `contains`, `isEmpty`, `isTrue`.
+
+### 6. Webhook (n8n-nodes-base.webhook)
+
+- **Description**: Receive external HTTP requests.
+- **Key Parameters**:
+  - `httpMethod`: GET, POST, etc.
+  - `path`: URL path.
+  - `responseMode`: "onReceived", "lastNode".
+
+### 7. Google Sheets (n8n-nodes-base.googleSheets)
+
+- **Description**: Read/Write data to Google Sheets.
+- **Key Parameters**:
+  - `resource`: "sheet".
+  - `operation`: "append", "update", "read".
+  - `sheetId`: Spreadsheet ID.
+  - `range`: Sheet name and range (e.g., "Sheet1!A:Z").
+
+### 8. Code (n8n-nodes-base.code)
+
+- **Description**: Run arbitrary JavaScript/TypeScript.
+- **Key Parameters**:
+  - `language`: "javaScript" or "typeScript".
+  - `jsCode`: The code string.
+
+### 9. Schedule Trigger (n8n-nodes-base.scheduleTrigger)
+
+- **Description**: Triggers a workflow on a time-based schedule (cron).
+- **Key Parameters**:
+  - `rule`: Object with `interval` array. Each entry has:
+    - `field`: `"cronExpression"` | `"hours"` | `"days"` | `"weeks"` | `"months"`
+    - `expression`: cron string when `field` is `"cronExpression"` (e.g. `"0 9 * * 1-5"`)
+    - `hoursInterval` / `minutesInterval`: numeric intervals for simpler schedules
+- **Notes**: Use `cronExpression` for precise schedules (e.g. every weekday at 9 AM). Do NOT use `n8n-nodes-base.cron` — it is deprecated.
+
+### 10. Execute Workflow (n8n-nodes-base.executeWorkflow)
+
+- **Description**: Call another n8n workflow as a sub-workflow and get its output.
+- **Key Parameters**:
+  - `source`: `"database"` (reference by ID) or `"localFile"` (path to JSON).
+  - `workflowId`: ID of the workflow to call (when `source` is `"database"`).
+  - `options.waitForSubWorkflow`: `true` to wait for the sub-workflow to finish (default).
+
+### 11. Execute Workflow Trigger (n8n-nodes-base.executeWorkflowTrigger)
+
+- **Description**: The trigger node that receives data when another workflow calls this one via Execute Workflow.
+- **Parameters**: None required. The node receives whatever data the calling workflow sends.
+- **Notes**: Use this as the start node in sub-workflows. Do NOT pair it with `n8n-nodes-base.start` in the same workflow.
+
+### 12. Merge (n8n-nodes-base.merge)
+
+- **Description**: Combine data from multiple input branches.
+- **Key Parameters**:
+  - `mode`: `"append"` | `"mergeByIndex"` | `"mergeByKey"` | `"multiplex"` | `"passThrough"` | `"wait"`
+  - `propertyName1` / `propertyName2`: Key fields for `"mergeByKey"` mode.
+- **Common use**: `"append"` to collect results from parallel branches; `"mergeByKey"` to join datasets on a shared field.
+
+### 13. Switch (n8n-nodes-base.switch)
+
+- **Description**: Route items to different outputs based on a value (multi-branch IF).
+- **Key Parameters**:
+  - `dataType`: `"string"` | `"number"` | `"boolean"`
+  - `value1`: The value to compare (expression).
+  - `rules.rules`: Array of `{ value2, outputKey }` pairs.
+  - `fallbackOutput`: Index to route non-matching items (default: ignored).
+
+### 14. Loop Over Items (n8n-nodes-base.splitInBatches)
+
+- **Description**: Iterate over a list in chunks, re-running downstream nodes for each batch.
+- **Key Parameters**:
+  - `batchSize`: Number of items per loop iteration.
+  - `options.reset`: `true` to reset the counter on the first run.
+- **Notes**: Connect the "loop" output back to the nodes you want to repeat. Connect the "done" output to what runs after the loop.
+
+### 15. Wait (n8n-nodes-base.wait)
+
+- **Description**: Pause execution and resume on a webhook call, a time delay, or a specific date/time.
+- **Key Parameters**:
+  - `resume`: `"timeInterval"` | `"specificTime"` | `"webhook"`
+  - `amount` + `unit`: For `"timeInterval"` (e.g. `amount: 5`, `unit: "minutes"`).
+  - `dateTime`: ISO string for `"specificTime"`.
+
+### 16. Respond to Webhook (n8n-nodes-base.respondToWebhook)
+
+- **Description**: Send an HTTP response back to the caller of a Webhook node.
+- **Key Parameters**:
+  - `respondWith`: `"text"` | `"json"` | `"binary"` | `"noData"` | `"redirect"`
+  - `responseBody`: The body to return (for `"text"` or `"json"`).
+  - `responseCode`: HTTP status code (default `200`).
+- **Notes**: Only valid when the upstream Webhook node has `responseMode: "lastNode"`. Must be the final node in the chain.
+
+### 17. Gmail (n8n-nodes-base.gmail)
+
+- **Description**: Send, receive, and manage Gmail messages and labels.
+- **Resources & Operations**:
+  - `message`: `send` | `get` | `getAll` | `delete` | `reply` | `addLabels` | `removeLabels` | `markAsRead` | `markAsUnread`
+  - `label`: `create` | `delete` | `get` | `getAll`
+- **Core send parameters**:
+  - `sendTo`: Recipient email address.
+  - `subject`: Email subject.
+  - `message`: Body text (HTML supported when `emailType: "html"`).
+  - `options.attachmentsUi`: List of binary field names to attach.
+
+### 18. Gmail Trigger (n8n-nodes-base.gmailTrigger)
+
+- **Description**: Polls Gmail and triggers when new messages arrive matching a filter.
+- **Key Parameters**:
+  - `filters.labelIds`: Array of Gmail label IDs to watch (e.g. `["INBOX", "UNREAD"]`).
+  - `pollTime.mode`: `"everyMinute"` | `"everyHour"` | `"custom"` (cron).
+  - `simple`: `true` to return simplified output; `false` for raw Gmail API response.
+
+### 19. Postgres (n8n-nodes-base.postgres)
+
+- **Description**: Execute queries against a PostgreSQL database.
+- **Key Parameters**:
+  - `operation`: `"executeQuery"` | `"insert"` | `"update"` | `"delete"` | `"select"`
+  - `query`: Raw SQL (for `"executeQuery"`).
+  - `table`: Target table name.
+  - `schema`: Schema name (default `"public"`).
+  - `columns`: Comma-separated column names for insert/update.
+- **Authentication**: Requires `postgres` credential type.
+
+### 20. MySQL (n8n-nodes-base.mySql)
+
+- **Description**: Execute queries against a MySQL/MariaDB database.
+- **Key Parameters**: Same pattern as Postgres (`operation`, `query`, `table`).
+- **Authentication**: Requires `mySql` credential type.
+
+### 21. Airtable (n8n-nodes-base.airtable)
+
+- **Description**: Read, create, update, and delete Airtable records.
+- **Resources & Operations**:
+  - `record`: `create` | `delete` | `get` | `list` | `update`
+- **Key Parameters**:
+  - `baseId`: Airtable Base ID (starts with `app`).
+  - `tableId`: Table ID or name.
+  - `fields`: Object of field values for create/update.
+  - `filterByFormula`: Airtable formula string for list filtering.
+
+### 22. Discord (n8n-nodes-base.discord)
+
+- **Description**: Send messages and manage content in Discord channels.
+- **Resources & Operations**:
+  - `message`: `send` | `get` | `getAll` | `delete` | `react` | `pin` | `unpin`
+  - `channel`: `create` | `delete` | `get` | `getAll` | `update`
+  - `member`: `ban` | `get` | `getAll` | `kick` | `roleAdd` | `roleRemove`
+- **Core send parameters**:
+  - `channelId`: Target channel ID.
+  - `content`: Message text. Supports Discord markdown.
+  - `options.embeds`: Array of embed objects.
+
+### 23. GitHub (n8n-nodes-base.github)
+
+- **Description**: Interact with the GitHub REST API — repos, issues, PRs, releases, files.
+- **Resources & Operations**:
+  - `issue`: `create` | `edit` | `get` | `getAll` | `lock` | `createComment`
+  - `pullRequest`: `create` | `get` | `getAll` | `createReview` | `merge`
+  - `release`: `create` | `delete` | `get` | `getAll` | `update`
+  - `file`: `create` | `delete` | `edit` | `get` | `list`
+  - `repository`: `get` | `getLicense` | `getProfile` | `listPopularPaths`
+- **Key Parameters**:
+  - `owner`: Repository owner (user or org).
+  - `repository`: Repository name.
+
+### 24. Notion (n8n-nodes-base.notion)
+
+- **Description**: Read and write Notion pages, databases, blocks, and users.
+- **Resources & Operations**:
+  - `database`: `get` | `getAll` | `search`
+  - `databasePage`: `create` | `get` | `getAll` | `update`
+  - `page`: `archive` | `create` | `get` | `search` | `update`
+  - `block`: `append` | `getAll` | `delete`
+  - `user`: `get` | `getAll`
+- **Key Parameters**:
+  - `pageId` / `databaseId`: Notion object IDs (32-char hex, no dashes required).
+  - `propertiesUi`: Collection of property values for page create/update.
+
+### 25. OpenAI (n8n-nodes-base.openAi)
+
+- **Description**: Call OpenAI APIs for text, image, audio, and embeddings.
+- **Resources & Operations**:
+  - `text`: `message` (chat completions)
+  - `image`: `generate` | `analyze`
+  - `audio`: `transcribe` | `translate` | `generateSpeech`
+  - `assistant`: `create` | `delete` | `get` | `list` | `message` | `update`
+  - `file`: `delete` | `get` | `list` | `upload`
+- **Key chat parameters**:
+  - `modelId`: e.g. `"gpt-4o"`, `"gpt-4o-mini"`.
+  - `messages.values`: Array of `{ role, content }` objects.
+  - `options.temperature` / `options.maxTokens`.
+
+### 26. AI Agent (n8n-nodes-langchain.agent)
+
+- **Description**: An LLM-powered agent that can call tools in a loop to complete a goal.
+- **Key Parameters**:
+  - `text`: The user prompt / task description. Supports `={{ $json.query }}`.
+  - `options.systemMessage`: System prompt to set the agent's persona.
+  - `options.maxIterations`: Maximum tool-call iterations (default 10).
+- **Notes**: This is a LangChain node (`n8n-nodes-langchain`), not `n8n-nodes-base`. It requires sub-nodes connected to the `ai_tool`, `ai_memory`, and `ai_languageModel` inputs.
+
+### 27. HTTP Request Tool (n8n-nodes-langchain.toolHttpRequest)
+
+- **Description**: Expose an HTTP endpoint as a tool available to an AI Agent node.
+- **Key Parameters**:
+  - `url`: The endpoint URL.
+  - `method`: HTTP method.
+  - `sendBody` / `bodyParameters`: Body to send.
+  - `description`: Human-readable description the LLM uses to decide when to call this tool.
+- **Notes**: Connect to the `ai_tool` input of an AI Agent node.
+
+### 28. Respond to AI Agent (n8n-nodes-langchain.toolWorkflow)
+
+- **Description**: Expose another n8n workflow as a callable tool for an AI Agent.
+- **Key Parameters**:
+  - `workflowId`: ID of the sub-workflow to invoke.
+  - `description`: What this tool does (seen by the LLM).
+
+---
+
+## Configuration Patterns
+
+- **Authentication**: Most nodes use either `authentication: "none"` or a
+  credential type string in `nodeCredentialType`.
+- **Expressions**: Use `={{ ... }}` for dynamic values, e.g.,
+  `={{ $json.fieldName }}`. Guard potentially null values:
+  `={{ $json.field ?? 'default' }}`.
+- **Accessing upstream nodes**: `$('Node Name').item.json.field` or
+  `$('Node Name').all()` to get all items from a named node.
+- **Binary data**: Access binary fields via `$binary.fieldName`. When passing
+  binary between nodes, ensure the field name matches exactly what the
+  producing node outputs (HTTP Request → `data`, Code nodes → whatever the
+  code sets).
+- **Node Connections**: Nodes are connected using the `connections` property in
+  the workflow JSON, mapping source node outputs to target node inputs.
+- **Error handling**: Add `"onError": "continueErrorOutput"` to a node's
+  settings to route errors to a secondary output instead of stopping the
+  workflow.
+- **typeVersion**: Always set the correct `typeVersion` for each node type.
+  When in doubt, use the version from the live n8n instance's node schema.
+  Using an incorrect version can cause silent failures or missing parameters.

package/docs/patterns/bigquery-via-http.md

@@ -0,0 +1,110 @@
+<!-- keywords: bigquery, google bigquery, bq, sql, analytics, data warehouse, dataset, table, query, merge, insert, drop, create table -->
+
+# Pattern: BigQuery Operations via HTTP Request
+
+## Critical Rule
+**NEVER use `n8n-nodes-base.googleBigQuery`** — it returns no output items for DDL/DML statements (CREATE, MERGE, UPDATE, DROP), breaking the workflow chain. Always use `n8n-nodes-base.httpRequest` with the BigQuery REST API instead.
+
+## Authentication
+All BigQuery HTTP nodes must use:
+```json
+{
+  "authentication": "predefinedCredentialType",
+  "nodeCredentialType": "googleApi"
+}
+```
+The `googleApi` credential must have scopes: `https://www.googleapis.com/auth/bigquery`
+
+---
+
+## DDL / DML Queries (CREATE, MERGE, UPDATE, DROP)
+
+Use the `jobs.query` synchronous endpoint. It always returns a JSON response, ensuring the workflow continues.
+
+```json
+{
+  "parameters": {
+    "method": "POST",
+    "url": "=https://bigquery.googleapis.com/bigquery/v2/projects/YOUR_PROJECT/queries",
+    "authentication": "predefinedCredentialType",
+    "nodeCredentialType": "googleApi",
+    "sendBody": true,
+    "specifyBody": "json",
+    "jsonBody": "={{ JSON.stringify({ query: 'YOUR SQL HERE', useLegacySql: false, timeoutMs: 30000 }) }}",
+    "options": {}
+  },
+  "type": "n8n-nodes-base.httpRequest",
+  "typeVersion": 4
+}
+```
+
+### CREATE TABLE example
+```json
+"jsonBody": "={{ JSON.stringify({ query: 'CREATE OR REPLACE TABLE `project.dataset.table` (Email STRING, Name STRING, Stripe_plan STRING, Unsubscribed BOOL)', useLegacySql: false, timeoutMs: 30000 }) }}"
+```
+
+### MERGE example (reference earlier node for dynamic table name)
+```json
+"jsonBody": "={{ JSON.stringify({ query: 'MERGE INTO `project.dataset.target` T USING `project.dataset.' + $('Parse CSV').item.json.stagingTable + '` S ON T.Email = S.Email WHEN MATCHED THEN UPDATE SET T.Name = S.Name WHEN NOT MATCHED THEN INSERT (Email, Name) VALUES (S.Email, S.Name)', useLegacySql: false, timeoutMs: 30000 }) }}"
+```
+
+### UPDATE example
+```json
+"jsonBody": "={{ JSON.stringify({ query: 'UPDATE `project.dataset.table` T SET T.active = false WHERE T.id NOT IN (SELECT id FROM `project.dataset.' + $('upstream').item.json.stagingTable + '`)', useLegacySql: false, timeoutMs: 30000 }) }}"
+```
+
+### DROP TABLE example
+```json
+"jsonBody": "={{ JSON.stringify({ query: 'DROP TABLE IF EXISTS `project.dataset.' + $('upstream').item.json.stagingTable + '`', useLegacySql: false, timeoutMs: 30000 }) }}"
+```
+
+---
+
+## Streaming Insert (insertAll)
+
+For inserting rows, use the `tabledata.insertAll` endpoint:
+
+```json
+{
+  "parameters": {
+    "method": "POST",
+    "url": "=https://bigquery.googleapis.com/bigquery/v2/projects/YOUR_PROJECT/datasets/YOUR_DATASET/tables/{{ $('upstream').item.json.stagingTable }}/insertAll",
+    "authentication": "predefinedCredentialType",
+    "nodeCredentialType": "googleApi",
+    "sendBody": true,
+    "specifyBody": "json",
+    "jsonBody": "={{ JSON.stringify({ rows: $('Parse CSV').item.json.rows }) }}",
+    "options": {}
+  },
+  "type": "n8n-nodes-base.httpRequest",
+  "typeVersion": 4
+}
+```
+
+Each row in `rows` must have an `insertId` (string) and `json` object:
+```json
+{ "insertId": "0", "json": { "Email": "user@example.com", "Name": "User" } }
+```
+
+---
+
+## Error Cleanup Pattern
+
+When a workflow creates a staging table, all intermediate nodes must have an error connection to a cleanup node that drops the table. This prevents orphaned tables on failure.
+
+```json
+"connections": {
+  "Create Staging Table": {
+    "main": [ [ { "node": "Insert Rows", "type": "main", "index": 0 } ] ],
+    "error": [ [ { "node": "Drop Staging Table (Error)", "type": "main", "index": 0 } ] ]
+  }
+}
+```
+
+The error cleanup node uses `$('Generate Table Name').item.json.stagingTable` (not `$('Parse CSV')`) because errors may fire before the CSV parsing node runs.
+
+---
+
+## Required IAM Roles for Service Account
+- `roles/bigquery.jobUser` — run jobs
+- `roles/bigquery.dataEditor` — read/write table data

package/oclif.manifest.json

@@ -122,13 +122,14 @@
       "aliases": [],
       "args": {
         "workflow": {
-          "description": "Path to the workflow file or workflow ID",
+          "description": "Path to workflow JSON file (omit for interactive menu)",
           "name": "workflow",
-          "required": true
+          "required": false
         }
       },
       "description": "Push workflows to n8n instance via API",
       "examples": [
+        "<%= config.bin %> <%= command.id %>",
         "<%= config.bin %> <%= command.id %> ./workflows/slack-notifier.json"
       ],
       "flags": {
@@ -147,6 +148,14 @@
           "name": "activate",
           "allowNo": false,
           "type": "boolean"
+        },
+        "dir": {
+          "char": "d",
+          "description": "Directory to scan for workflows (default: ./workflows)",
+          "name": "dir",
+          "hasDynamicHelp": false,
+          "multiple": false,
+          "type": "option"
         }
       },
       "hasDynamicHelp": false,
@@ -239,6 +248,76 @@
         "fixture.js"
       ]
     },
+    "learn": {
+      "aliases": [],
+      "args": {
+        "workflow": {
+          "description": "Path to workflow JSON file (omit for interactive menu)",
+          "name": "workflow",
+          "required": false
+        }
+      },
+      "description": "Extract reusable patterns from a validated workflow into the pattern library",
+      "examples": [
+        "<%= config.bin %> <%= command.id %>",
+        "<%= config.bin %> <%= command.id %> ./workflows/my-workflow/workflow.json",
+        "<%= config.bin %> <%= command.id %> --github owner/repo",
+        "<%= config.bin %> <%= command.id %> --github owner/repo --github-path patterns/google"
+      ],
+      "flags": {
+        "dir": {
+          "char": "d",
+          "description": "Directory to scan for workflows (default: ./workflows)",
+          "name": "dir",
+          "hasDynamicHelp": false,
+          "multiple": false,
+          "type": "option"
+        },
+        "all": {
+          "description": "Generate patterns for all workflows in the directory",
+          "name": "all",
+          "allowNo": false,
+          "type": "boolean"
+        },
+        "github": {
+          "description": "Import patterns from a GitHub repo (format: owner/repo or owner/repo@branch)",
+          "name": "github",
+          "hasDynamicHelp": false,
+          "multiple": false,
+          "type": "option"
+        },
+        "github-path": {
+          "description": "Path within the GitHub repo to import from (default: patterns)",
+          "name": "github-path",
+          "default": "patterns",
+          "hasDynamicHelp": false,
+          "multiple": false,
+          "type": "option"
+        },
+        "token": {
+          "description": "GitHub personal access token (increases rate limit for public repos)",
+          "env": "GITHUB_TOKEN",
+          "name": "token",
+          "hasDynamicHelp": false,
+          "multiple": false,
+          "type": "option"
+        }
+      },
+      "hasDynamicHelp": false,
+      "hiddenAliases": [],
+      "id": "learn",
+      "pluginAlias": "@lhi/n8m",
+      "pluginName": "@lhi/n8m",
+      "pluginType": "core",
+      "strict": true,
+      "enableJsonFlag": false,
+      "isESM": true,
+      "relativePath": [
+        "dist",
+        "commands",
+        "learn.js"
+      ]
+    },
     "mcp": {
       "aliases": [],
       "args": {},
@@ -439,5 +518,5 @@
       ]
     }
   },
-  "version": "0.2.3"
+  "version": "0.3.0"
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lhi/n8m",
-  "version": "0.2.3",
+  "version": "0.3.0",
   "description": "Agentic n8n CLI wrapper - A Skill Bridge for n8n workflow automation",
   "author": "Lem Canady",
   "license": "MIT",
@@ -15,6 +15,7 @@
   "files": [
     "bin",
     "dist",
+    "docs",
     "oclif.manifest.json"
   ],
   "dependencies": {
@@ -94,6 +95,7 @@
     "n8m": "./bin/run.js",
     "start": "npm run build && ./bin/run.js",
     "dev": "tsc -b -w",
+    "generate-patterns": "tsx scripts/generate-patterns.ts",
     "version": "oclif readme && git add README.md"
   }
 }