# Glean

> Search your company's knowledge with Glean, ask Glean Assistant and get answers with their sources, summarize documents, run Glean agents, look people up, and manage answers, collections, go links and announcements.

App · 26 actions. Web page: https://docs.lodol.com/docs/integrations/glean

## Actions

### Add item to collection

Add a link to a collection, or a text note when no address is given. Returns the new item and its ID.

**Inputs**

- Collection (`collection_id`, text, required): The collection. Pick one, or give its number from an earlier step.
- Address (`url`, text): The web address to add, such as a document's url.
- Name (`name`, text): A name to show for it. Needed for a text note.
- Description (`description`, text): Why it is in the collection.

**Outputs**

- `collection` (object): A collection
- `item` (object): An item in a collection
- `item_id` (text): The new item's ID, for 'Remove item from collection'

**Example output**

```json
{
  "collection": {
    "created_at": "2026-10-24T09:00:00Z",
    "creator": "Jane Doe",
    "description": "Everything a new hire needs.",
    "icon": "\ud83d\udcda",
    "id": 42,
    "item_count": 1,
    "items": [
      {
        "description": "Where we keep runbooks.",
        "document_id": "",
        "item_id": "item_1",
        "item_type": "URL",
        "name": "Team wiki",
        "url": "https://example.com/wiki"
      }
    ],
    "name": "Onboarding",
    "updated_at": "2026-10-24T09:00:00Z"
  },
  "item": {
    "description": "Where we keep runbooks.",
    "document_id": "",
    "item_id": "item_1",
    "item_type": "URL",
    "name": "Team wiki",
    "url": "https://example.com/wiki"
  },
  "item_id": "item_1"
}
```

### Ask Glean Assistant

Ask Glean Assistant a question about your company's knowledge. Returns the answer as plain text, the sources it cites, and a conversation ID for a follow-up question.

**Inputs**

- Question (`question`, text, required): What to ask, in your own words.
- Continue conversation (`conversation_id`, text): A conversation ID from an earlier answer, to ask a follow-up. Leave empty to start a new conversation.

**Outputs**

- `answer` (text): The answer, as plain text
- `citations` (list): The sources it cites
- `conversation_id` (text): Give this to ask a follow-up

**Example output**

```json
{
  "answer": "Book travel through Navan; flights over 6 hours may be business class.",
  "citations": [
    {
      "datasource": "gdrive",
      "title": "Travel policy",
      "type": "document",
      "url": "https://example.com/travel"
    }
  ],
  "conversation_id": "conv_01HX"
}
```

### Create announcement

Post an announcement that everyone at your company sees in Glean between a start and an end time. You need permission to post announcements in Glean.

**Inputs**

- Title (`title`, text, required): The headline.
- Text (`body`, text): The announcement's text.
- Starts (`starts_at`, date and time): When it starts showing. Leave empty for now.
- Ends (`ends_at`, date and time, required): When it stops showing.

**Outputs**

- `announcement` (object): An announcement

**Example output**

```json
{
  "announcement": {
    "author": "Jane Doe",
    "body": "The office is closed for the holiday.",
    "created_at": "2026-10-24T09:00:00Z",
    "ends_at": "2026-10-31T17:00:00Z",
    "id": 11,
    "is_published": true,
    "starts_at": "2026-10-27T09:00:00Z",
    "title": "Office closed Friday",
    "view_url": ""
  }
}
```

### Create answer

Add a question and its answer to Glean, so search and Glean Assistant can give it to others.

**Inputs**

- Question (`question`, text, required): The question people ask.
- Answer (`answer_text`, text, required): The answer, as plain text.
- Other ways to ask (`question_variations`, list): Other phrasings of the question, so more searches find it.

**Outputs**

- `answer` (object): An answer

**Example output**

```json
{
  "answer": {
    "answer_text": "Blue light is scattered more than red light.",
    "author": "Jane Doe",
    "created_at": "2026-10-24T09:00:00Z",
    "doc_id": "ANSWERS_answer_3",
    "id": 3,
    "question": "Why is the sky blue?",
    "question_variations": [],
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Create collection

Create a collection to gather documents and links under one name.

**Inputs**

- Name (`name`, text, required): The collection's name, which must be unique.
- Description (`description`, text): What the collection is for.
- Emoji (`icon`, text): An emoji to show with it, such as 📚.

**Outputs**

- `collection` (object): A collection

**Example output**

```json
{
  "collection": {
    "created_at": "2026-10-24T09:00:00Z",
    "creator": "Jane Doe",
    "description": "Everything a new hire needs.",
    "icon": "\ud83d\udcda",
    "id": 42,
    "item_count": 0,
    "items": [],
    "name": "Onboarding",
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Create go link

Create a go link, such as go/team-wiki, that takes everyone at your company to an address.

**Inputs**

- Name (`alias`, text, required): The part after go/, such as team-wiki.
- Goes to (`destination_url`, text, required): The web address the go link opens.
- Description (`description`, text): What the go link is for.
- Unlisted (`unlisted`, true/false): Hide it from everyone but you and Glean admins.

**Outputs**

- `go_link` (object): A go link

**Example output**

```json
{
  "go_link": {
    "alias": "team-wiki",
    "created_at": "2026-10-24T09:00:00Z",
    "created_by": "Jane Doe",
    "description": "The team wiki.",
    "destination_url": "https://example.com/wiki",
    "id": 7,
    "link": "go/team-wiki",
    "unlisted": false,
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Delete announcement

Delete an announcement, taking it down if it is showing.

**Inputs**

- Announcement (`announcement_id`, text, required): The announcement's number, from 'Create announcement'.

**Outputs**

- `announcement_id` (number): The announcement deleted
- `deleted` (true/false): Always true once deleted

**Example output**

```json
{
  "announcement_id": 11,
  "deleted": true
}
```

### Delete answer

Delete an answer from Glean.

**Inputs**

- Answer (`answer_id`, text, required): The answer's number, from 'Create answer' or an earlier step.

**Outputs**

- `answer_id` (text): The answer deleted
- `deleted` (true/false): Always true once deleted

**Example output**

```json
{
  "answer_id": "3",
  "deleted": true
}
```

### Delete collection

Delete a collection. The documents in it are not touched.

**Inputs**

- Collection (`collection_id`, text, required): The collection. Pick one, or give its number from an earlier step.

**Outputs**

- `collection_id` (number): The collection deleted
- `deleted` (true/false): Always true once deleted

**Example output**

```json
{
  "collection_id": 42,
  "deleted": true
}
```

### Delete go link

Delete a go link, so it no longer opens anything.

**Inputs**

- Go link (`go_link`, text, required): The go link. Pick one, or give its name (such as team-wiki) or number.

**Outputs**

- `deleted` (true/false): Always true once deleted
- `go_link_id` (number): The go link deleted

**Example output**

```json
{
  "deleted": true,
  "go_link_id": 7
}
```

### Find people

Find people in your company directory by name, title, team or other details.

**Inputs**

- Search for (`query`, text): A name, title, team or other detail. Leave empty to list everyone.
- Page size (`page_size`, number): How many to return, 1 to 100. From 1 to 100.
- Next page (`cursor`, text): The 'next_cursor' from an earlier page, to get the page after it.

**Outputs**

- `count` (number): How many on this page
- `has_more` (true/false): True if there are more
- `next_cursor` (text): Give as 'Next page' for the page after; empty on the last
- `people` (list): The people found

**Example output**

```json
{
  "count": 1,
  "has_more": false,
  "next_cursor": "",
  "people": [
    {
      "business_unit": "Product",
      "department": "Engineering",
      "email": "jane@example.com",
      "employee_type": "FULL_TIME",
      "id": "abc123",
      "location": "San Francisco",
      "manager_email": "sam@example.com",
      "manager_id": "def456",
      "manager_name": "Sam Lee",
      "name": "Jane Doe",
      "phone": "",
      "photo_url": "",
      "start_date": "2022-03-01",
      "timezone": "America/Los_Angeles",
      "title": "Engineering Manager"
    }
  ]
}
```

### Get agent run

Check on an agent run started by 'Run agent': its state and what the agent has said.

**Inputs**

- Agent (`agent_id`, text, required): The agent. Pick one, or give its ID from 'List agents' or the end of its address in Agent Builder.
- Run ID (`run_id`, text, required): The run_id from 'Run agent'.

**Outputs**

- `agent_id` (text): The agent's ID
- `answer` (text): What the agent said, as plain text
- `created_at` (text): When the run started (ISO 8601)
- `error` (text): Why the run failed, if it did
- `finished` (true/false): True once the run has ended
- `messages` (list): The run's messages in order
- `run_id` (text): The run's ID, for 'Get agent run'
- `state` (text): QUEUED, RUNNING, REQUIRES_INPUT, SUCCEEDED, FAILED, CANCELLING, CANCELLED or EXPIRED
- `updated_at` (text): When the run last changed (ISO 8601)
- `waiting_for_approval` (list): Tools the agent wants to use that someone must approve in Glean

**Example output**

```json
{
  "agent_id": "mho4lwzylcozgoc2",
  "answer": "Parental leave is 16 weeks at full pay.",
  "created_at": "2026-10-24T09:00:00Z",
  "error": "",
  "finished": true,
  "messages": [
    {
      "role": "USER",
      "text": "What is our parental leave policy?"
    },
    {
      "role": "GLEAN_AI",
      "text": "Parental leave is 16 weeks at full pay."
    }
  ],
  "run_id": "run_01JAGENT",
  "state": "SUCCEEDED",
  "updated_at": "2026-10-24T09:00:20Z",
  "waiting_for_approval": []
}
```

### Get answer

Get one answer: its question, answer, author and dates.

**Inputs**

- Answer (`answer_id`, text, required): The answer's number, from 'Create answer' or an earlier step.

**Outputs**

- `answer` (object): An answer

**Example output**

```json
{
  "answer": {
    "answer_text": "Blue light is scattered more than red light.",
    "author": "Jane Doe",
    "created_at": "2026-10-24T09:00:00Z",
    "doc_id": "ANSWERS_answer_3",
    "id": 3,
    "question": "Why is the sky blue?",
    "question_variations": [],
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Get collection

Get one collection with every item in it.

**Inputs**

- Collection (`collection_id`, text, required): The collection. Pick one, or give its number from an earlier step.

**Outputs**

- `collection` (object): A collection

**Example output**

```json
{
  "collection": {
    "created_at": "2026-10-24T09:00:00Z",
    "creator": "Jane Doe",
    "description": "Everything a new hire needs.",
    "icon": "\ud83d\udcda",
    "id": 42,
    "item_count": 1,
    "items": [
      {
        "description": "Where we keep runbooks.",
        "document_id": "",
        "item_id": "item_1",
        "item_type": "URL",
        "name": "Team wiki",
        "url": "https://example.com/wiki"
      }
    ],
    "name": "Onboarding",
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Get documents

Get documents' details from Glean (title, app, author, owner, dates) and, if you want, their full text.

**Inputs**

- Documents (`documents`, list, required): Up to 20 documents, each by its web address (such as a search result's url) or its Glean document ID.
- Include full text (`include_content`, true/false): Also return each document's full text.

**Outputs**

- `count` (number): How many were found
- `documents` (list): The documents found
- `not_found` (list): Documents Glean could not find or you can't open

**Example output**

```json
{
  "count": 1,
  "documents": [
    {
      "author": "Jane Doe",
      "container": "Engineering",
      "content": "",
      "created_at": "2026-04-01T09:00:00Z",
      "datasource": "confluence",
      "doc_type": "page",
      "id": "CONFLUENCE_12345",
      "owner": "Jane Doe",
      "status": "",
      "title": "Q2 2026 Platform Roadmap",
      "updated_at": "2026-10-20T14:00:00Z",
      "url": "https://company.atlassian.net/wiki/spaces/ENG/pages/12345"
    }
  ],
  "not_found": []
}
```

### Get go link

Get one go link: where it leads, who made it and when.

**Inputs**

- Go link (`go_link`, text, required): The go link. Pick one, or give its name (such as team-wiki) or number.

**Outputs**

- `go_link` (object): A go link

**Example output**

```json
{
  "go_link": {
    "alias": "team-wiki",
    "created_at": "2026-10-24T09:00:00Z",
    "created_by": "Jane Doe",
    "description": "The team wiki.",
    "destination_url": "https://example.com/wiki",
    "id": 7,
    "link": "go/team-wiki",
    "unlisted": false,
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Get person

Get someone's profile from your company directory: title, department, location, manager and more. Leave the person empty to get your own.

**Inputs**

- Email or person ID (`person`, text): Their work email, or a person ID from an earlier step. Leave empty for yourself.

**Outputs**

- `person` (object): A person

**Example output**

```json
{
  "person": {
    "business_unit": "Product",
    "department": "Engineering",
    "email": "jane@example.com",
    "employee_type": "FULL_TIME",
    "id": "abc123",
    "location": "San Francisco",
    "manager_email": "sam@example.com",
    "manager_id": "def456",
    "manager_name": "Sam Lee",
    "name": "Jane Doe",
    "phone": "",
    "photo_url": "",
    "start_date": "2022-03-01",
    "timezone": "America/Los_Angeles",
    "title": "Engineering Manager"
  }
}
```

### List agents

List the Glean agents you can run, optionally only those whose name contains some text.

**Inputs**

- Name contains (`name`, text): Only agents whose name contains this. Leave empty for all.

**Outputs**

- `agents` (list): The agents
- `count` (number): How many

**Example output**

```json
{
  "agents": [
    {
      "agent_id": "mho4lwzylcozgoc2",
      "description": "Answers HR policy questions.",
      "name": "HR Policy Agent"
    }
  ],
  "count": 1
}
```

### List collections

List the collections you can see, optionally only those whose name contains some text.

**Inputs**

- Name contains (`name`, text): Only collections whose name contains this. Leave empty for all.

**Outputs**

- `collections` (list): The collections
- `count` (number): How many

**Example output**

```json
{
  "collections": [
    {
      "created_at": "2026-10-24T09:00:00Z",
      "creator": "Jane Doe",
      "description": "Everything a new hire needs.",
      "icon": "\ud83d\udcda",
      "id": 42,
      "item_count": 1,
      "items": [],
      "name": "Onboarding",
      "updated_at": "2026-10-24T09:00:00Z"
    }
  ],
  "count": 1
}
```

### List go links

List the go links you can edit, optionally only those matching some text.

**Inputs**

- Search for (`search`, text): Text in the name, address or description. Leave empty for all.
- Page size (`page_size`, number): How many to return, 1 to 100. From 1 to 100.
- Next page (`cursor`, text): The 'next_cursor' from an earlier page, to get the page after it.

**Outputs**

- `count` (number): How many on this page
- `go_links` (list): The go links
- `has_more` (true/false): True if there are more
- `next_cursor` (text): Give as 'Next page' for the page after; empty on the last
- `total` (number): How many match in all

**Example output**

```json
{
  "count": 1,
  "go_links": [
    {
      "alias": "team-wiki",
      "created_at": "2026-10-24T09:00:00Z",
      "created_by": "Jane Doe",
      "description": "The team wiki.",
      "destination_url": "https://example.com/wiki",
      "id": 7,
      "link": "go/team-wiki",
      "unlisted": false,
      "updated_at": "2026-10-24T09:00:00Z"
    }
  ],
  "has_more": false,
  "next_cursor": "",
  "total": 1
}
```

### Remove item from collection

Take one item out of a collection. The document itself is not touched.

**Inputs**

- Collection (`collection_id`, text, required): The collection. Pick one, or give its number from an earlier step.
- Item ID (`item_id`, text, required): The item's ID, from 'Get collection' or 'Add item to collection'.

**Outputs**

- `collection` (object): A collection
- `item_id` (text): The item removed
- `removed` (true/false): Always true once removed

**Example output**

```json
{
  "collection": {
    "created_at": "2026-10-24T09:00:00Z",
    "creator": "Jane Doe",
    "description": "Everything a new hire needs.",
    "icon": "\ud83d\udcda",
    "id": 42,
    "item_count": 0,
    "items": [],
    "name": "Onboarding",
    "updated_at": "2026-10-24T09:00:00Z"
  },
  "item_id": "item_1",
  "removed": true
}
```

### Run agent

Run a Glean agent with a message or its form's fields, and wait up to five minutes for its answer. A run that takes longer comes back unfinished with a run ID for 'Get agent run'.

**Inputs**

- Agent (`agent_id`, text, required): The agent. Pick one, or give its ID from 'List agents' or the end of its address in Agent Builder.
- Message (`message`, text): What to ask the agent, for an agent that takes a message.
- Form fields (`inputs`, object): The agent's form fields, for an agent with an input form. Fields left empty are sent empty.

**Outputs**

- `agent_id` (text): The agent's ID
- `answer` (text): What the agent said, as plain text
- `created_at` (text): When the run started (ISO 8601)
- `error` (text): Why the run failed, if it did
- `finished` (true/false): True once the run has ended
- `messages` (list): The run's messages in order
- `run_id` (text): The run's ID, for 'Get agent run'
- `state` (text): QUEUED, RUNNING, REQUIRES_INPUT, SUCCEEDED, FAILED, CANCELLING, CANCELLED or EXPIRED
- `updated_at` (text): When the run last changed (ISO 8601)
- `waiting_for_approval` (list): Tools the agent wants to use that someone must approve in Glean

**Example output**

```json
{
  "agent_id": "mho4lwzylcozgoc2",
  "answer": "Parental leave is 16 weeks at full pay.",
  "created_at": "2026-10-24T09:00:00Z",
  "error": "",
  "finished": true,
  "messages": [
    {
      "role": "USER",
      "text": "What is our parental leave policy?"
    },
    {
      "role": "GLEAN_AI",
      "text": "Parental leave is 16 weeks at full pay."
    }
  ],
  "run_id": "run_01JAGENT",
  "state": "SUCCEEDED",
  "updated_at": "2026-10-24T09:00:20Z",
  "waiting_for_approval": []
}
```

### Search

Search everything you can see in Glean, across all your company's connected apps. Returns one page of results with their titles, addresses and matching passages.

**Inputs**

- Search for (`query`, text, required): What to search for. Glean's operators work too, such as from:jane or app:confluence.
- Apps (`datasources`, list): Only results from these apps. Leave empty for all.
- Updated on or after (`updated_after`, date): Only results last changed on or after this date.
- Updated on or before (`updated_before`, date): Only results last changed on or before this date.
- Page size (`page_size`, number): How many to return, 1 to 100. From 1 to 100.
- Next page (`cursor`, text): The 'next_cursor' from an earlier page, to get the page after it.

**Outputs**

- `count` (number): How many on this page
- `has_more` (true/false): True if there are more
- `next_cursor` (text): Give as 'Next page' for the page after; empty on the last
- `results` (list): The results, best first
- `warnings` (list): Anything Glean noted, such as partial results

**Example output**

```json
{
  "count": 1,
  "has_more": false,
  "next_cursor": "",
  "results": [
    {
      "created_at": "2026-04-01T09:00:00Z",
      "creator": "Jane Doe",
      "datasource": "confluence",
      "document_type": "page",
      "owner": "Jane Doe",
      "snippets": [
        "The platform team will focus on API stability and..."
      ],
      "title": "Q2 2026 Platform Roadmap",
      "updated_at": "2026-10-20T14:00:00Z",
      "url": "https://company.atlassian.net/wiki/spaces/ENG/pages/12345"
    }
  ],
  "warnings": []
}
```

### Summarize documents

Have Glean write a short summary of one or more documents, optionally about a question you give.

**Inputs**

- Documents (`documents`, list, required): Up to 20 documents, each by its web address (such as a search result's url) or its Glean document ID.
- Focus on (`focus`, text): A question or topic the summary should be about.
- Length (`summary_length`, number): About how many characters, 100 to 5000. Glean's default is 500. From 100 to 5000.

**Outputs**

- `follow_up_prompts` (list): Questions worth asking next
- `summary` (text): The summary

**Example output**

```json
{
  "follow_up_prompts": [
    "Who owns API stability?"
  ],
  "summary": "The roadmap puts API stability first, then search speed."
}
```

### Update answer

Change an answer's question, answer or other phrasings. Only the fields given change.

**Inputs**

- Answer (`answer_id`, text, required): The answer's number, from 'Create answer' or an earlier step.
- Question (`question`, text): A new question. Leave empty to keep it.
- Answer (`answer_text`, text): A new answer. Leave empty to keep it.
- Other ways to ask (`question_variations`, list): Other phrasings of the question, so more searches find it.

**Outputs**

- `answer` (object): An answer

**Example output**

```json
{
  "answer": {
    "answer_text": "Rayleigh scattering.",
    "author": "Jane Doe",
    "created_at": "2026-10-24T09:00:00Z",
    "doc_id": "ANSWERS_answer_3",
    "id": 3,
    "question": "Why is the sky blue?",
    "question_variations": [],
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```

### Update go link

Change a go link's name, address, description or visibility. Only the fields given change.

**Inputs**

- Go link (`go_link`, text, required): The go link. Pick one, or give its name (such as team-wiki) or number.
- New name (`alias`, text): A new name after go/. Leave empty to keep it.
- Goes to (`destination_url`, text): A new address. Leave empty to keep it.
- Description (`description`, text): A new description. Leave empty to keep it.
- Unlisted (`unlisted`, true/false): Hide it from everyone but you and Glean admins.

**Outputs**

- `go_link` (object): A go link

**Example output**

```json
{
  "go_link": {
    "alias": "team-wiki",
    "created_at": "2026-10-24T09:00:00Z",
    "created_by": "Jane Doe",
    "description": "The team wiki.",
    "destination_url": "https://example.com/new",
    "id": 7,
    "link": "go/team-wiki",
    "unlisted": false,
    "updated_at": "2026-10-24T09:00:00Z"
  }
}
```
