### Install Agent Skills via CLI

Source: https://github.com/thinkcol/fasta-skills/blob/master/README.md

Use these commands to add FASTA skills to your project. Global installation or targeting specific agents is supported via flags.

```bash
# Install all skills
npx skills add ThinkCol/fasta-skills

# Install a specific skill
npx skills add ThinkCol/fasta-skills --skill fasta-adhocsearch-api
npx skills add ThinkCol/fasta-skills --skill open-lenx-api
```

--------------------------------

### Search via cURL

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/fasta-adhocsearch-api/SKILL.md

Example request using cURL with environment variable placeholders.

```bash
curl -X POST "${ADHOCSEARCH_ENDPOINT:-https://adhocsearch.fasta.ai}/search" \
  -H "Content-Type: application/json" \
  -H "x-api-key: $ADHOCSEARCH_API_KEY" \
  -H "x-user-id: $ADHOCSEARCH_USER_ID" \
  -d '{
    "query": "climate change",
    "from": 1711929600,
    "to": 1714521600,
    "size": 10
  }'
```

--------------------------------

### Get Task Data using curl

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Example using curl to retrieve task data within a specified date range and size. Ensure to set `LENX_ENDPOINT`, `LENX_API_KEY`, and `LENX_USER_ID` environment variables.

```bash
curl -s "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks/42/data?from=1711929600&to=1714521600&size=10" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID"
```

--------------------------------

### Search via Python Requests

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/fasta-adhocsearch-api/SKILL.md

Example request using the Python requests library.

```python
import os, requests

endpoint = os.getenv("ADHOCSEARCH_ENDPOINT", "https://adhocsearch.fasta.ai")
resp = requests.post(
    f"{endpoint}/search",
    headers={
        "Content-Type": "application/json",
        "x-api-key": os.environ["ADHOCSEARCH_API_KEY"],
        "x-user-id": os.environ["ADHOCSEARCH_USER_ID"],
    },
    json={
        "keywords": ["election", "vote"],
        "from": 1711929600,
        "to": 1714521600,
        "size": 50,
    },
)
data = resp.json()["data"]
```

--------------------------------

### Get Tasks

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

List all tasks accessible to the authenticated user, with pagination.

```APIDOC
## GET /api/v1/tasks

### Description
List all tasks accessible to the authenticated user, with pagination.

### Method
GET

### Endpoint
`{endpoint}/api/v1/tasks`

### Parameters
#### Query Parameters
- **page** (number) - Optional - Page number (default: 1)
- **size** (number) - Optional - Results per page, max 20 (default: 10)

### Response
#### Success Response (200 OK)
Cached 5 minutes. Returns:
```json
{
  "data": [
    {
      "task_id": 123,
      "task_name": "My Task",
      "task_type": 0,
      "indus_id": null,
      "labels": [],
      "lang_abbr": "en",
      "status": 1,
      "created_at": "2025-01-15T10:30:00.000Z",
      "updated_at": "2025-03-20T14:00:00.000Z"
    }
  ],
  "pagination": {
    "page": 1,
    "size": 10,
    "total": 42,
    "totalPages": 5
  }
}
```
Each item contains: `task_id`, `task_name`, `task_type`, `indus_id`, `labels`, `lang_abbr`, `status`, `created_at`, `updated_at`. Results are sorted by `updated_at` descending.

#### Errors
- **400** — Invalid query parameters
- **401** — Invalid API key or user ID
- **500** — Unexpected error
```

--------------------------------

### GET /api/v1/tasks

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

List all accessible monitoring tasks with pagination support.

```APIDOC
## GET /api/v1/tasks

### Description
List all accessible monitoring tasks. Results are sorted by updated_at descending.

### Method
GET

### Endpoint
/api/v1/tasks

### Parameters
#### Query Parameters
- **page** (integer) - Required - Page number.
- **size** (integer) - Required - Number of items per page (max 20).

### Response
#### Success Response (200)
- **data** (array) - List of task objects.
- **pagination** (object) - Pagination metadata including page, size, total, and totalPages.
```

--------------------------------

### GET /api/v1/tasks

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Lists tasks with pagination support.

```APIDOC
## GET /api/v1/tasks

### Description
Retrieves a paginated list of tasks, sorted by updated_at descending.

### Method
GET

### Endpoint
/api/v1/tasks

### Parameters
#### Query Parameters
- **page** (integer) - Optional - The page number to retrieve.
- **size** (integer) - Optional - The number of items per page (max 20).
```

--------------------------------

### Get Task Info

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Retrieve metadata and search configuration for a specific monitoring task.

```APIDOC
## GET /api/v1/tasks/{task_id}

### Description
Retrieve metadata and search configuration for a specific task.

### Method
GET

### Endpoint
`{endpoint}/api/v1/tasks/{task_id}`

### Parameters
#### Path Parameters
- **task_id** (number) - Required - monitoring task ID

### Response
#### Success Response (200 OK)
Cached 5 minutes. Returns `{ "data": { ... } }` with fields: `task_id`, `task_name`, `owner_ac_id`, `task_type`, `indus_id`, `labels`, `lang_abbr`, `status`, `region`, `created_at`, `updated_at`, `search_query` (object or null with: `lang_abbr`, `region`, `list_medium`, `list_author_id`, `query_layer`, `query_string`, `cron`, `exclude_channel_links`, `updated_at`)

#### Errors
- **400** — Invalid request
- **401** — Invalid API key or user ID
- **403** — User does not have permission to access this task
- **404** — Task not found
```

--------------------------------

### Validate query_layer Structures

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Examples of valid and invalid JSON structures for the query_layer.

```json
{ "in": ["a", "b"] }
{ "in": [["a", "b"], ["c"]] }
```

```json
{ "in": ["a", ["b"]] }
```

--------------------------------

### GET /api/v1/tasks/{taskId}

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Retrieve detailed metadata and search configuration for a specific monitoring task.

```APIDOC
## GET /api/v1/tasks/{taskId}

### Description
Retrieve detailed metadata and search configuration for a specific monitoring task. Response is cached for 5 minutes.

### Method
GET

### Endpoint
/api/v1/tasks/{taskId}

### Parameters
#### Path Parameters
- **taskId** (integer) - Required - The unique identifier of the task.

### Response
#### Success Response (200)
- **data** (object) - Task details including task_id, task_name, owner_ac_id, task_type, indus_id, labels, lang_abbr, status, region, created_at, updated_at, and search_query.
```

--------------------------------

### GET /api/v1/tasks/{id}/data

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Retrieves data for a specific task with pagination.

```APIDOC
## GET /api/v1/tasks/{id}/data

### Description
Retrieves data associated with a specific task within a defined time range.

### Method
GET

### Endpoint
/api/v1/tasks/{id}/data

### Parameters
#### Path Parameters
- **id** (string) - Required - The unique identifier of the task.

#### Query Parameters
- **from** (integer) - Required - Start Unix timestamp.
- **to** (integer) - Required - End Unix timestamp.
- **size** (integer) - Required - Number of records to return.
- **search_after** (string) - Optional - Cursor for pagination.

### Response
#### Success Response (200)
- **data** (array) - List of task data records.
```

--------------------------------

### Get Task Data

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Fetch monitoring data for a specific task within a given date range.

```APIDOC
## GET /api/v1/tasks/{task_id}/data

### Description
Fetch monitoring data for a specific task within a given date range.

### Method
GET

### Endpoint
`{endpoint}/api/v1/tasks/{task_id}/data`

### Parameters
#### Path Parameters
- **task_id** (number) - Required - monitoring task ID

#### Query Parameters
- **from** (number) - Required - Start of date range (Unix timestamp, seconds)
- **to** (number) - Required - End of date range (Unix timestamp, seconds)
- **size** (number) - Required - Results to return (1–1000)
- **search_after** (number) - Optional - Pagination cursor — `unix_timestamp` of last result from previous page

**Pagination:** Results are sorted by `unix_timestamp` descending. Pass the last item's `unix_timestamp` as `search_after` to get the next page. Repeat until fewer results than `size` are returned.

### Response
#### Success Response (200 OK)
Cached 5 minutes. Returns `{ "data": [...] }` where each item contains: `id`, `hash`, `country`, `lang_abbr`, `medium`, `channel`, `channel_link`, `site`, `thread_link`, `post_link`, `thread_title`, `post_message`, `post_timestamp` (ISO 8601), `unix_timestamp`, `author_name`, `author_id`, `author_image`, `author_link`, `is_comment`, `comment_order`, `comment_count`, `share_count`, `view_count`, `reaction_count`, `reaction_like`, `reaction_dislike`, `reaction_love`, `reaction_wow`, `reaction_haha`, `reaction_sad`, `reaction_angry`, `sentiment_score`

#### Errors
- **400** — `{ "message": "Invalid request body", "error": "..." }`
- **401** — Invalid API key or user ID
- **403** — `{ "message": "Forbidden", "error": "User does not have permission to access this task" }
```

--------------------------------

### GET /api/v1/tasks/{task_id}/data

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Retrieve monitoring data for a specific task within a date range, supporting pagination via search_after.

```APIDOC
## GET /api/v1/tasks/{task_id}/data

### Description
Retrieve monitoring data (social media posts) for a specific task within a date range. Supports pagination for large datasets using `search_after` with the last result's `unix_timestamp`.

### Method
GET

### Endpoint
/api/v1/tasks/{task_id}/data

### Parameters
#### Path Parameters
- **task_id** (integer) - Required - The ID of the task to retrieve data for.

#### Query Parameters
- **from** (integer) - Required - Start unix timestamp.
- **to** (integer) - Required - End unix timestamp.
- **size** (integer) - Optional - Number of results to return (max 1000).
- **search_after** (integer) - Optional - The unix_timestamp of the last result from the previous page.
```

--------------------------------

### Create Monitoring Task

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Initializes a new monitoring task using a POST request with a defined search query configuration.

```bash
# Create a live monitoring task for brand mentions
curl -s -X POST "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID" \
  -H "Content-Type: application/json" \
  -d '{
    "task_type": "live",
    "task_name": "Brand Monitoring - ACME Corp",
    "language": "en",
    "search_query": {
      "query_layer": [
        {
          "in": [["acme corp", "acme corporation"], ["product", "service"]],
          "ex": ["job listing", "careers"]
        }
      ],
      "region": "USA",
      "list_medium": ["News", "Social", "Facebook"]
    }
  }'

# Response (201 Created):
# {
#   "data": {
#     "task_id": 456,
#     "task_name": "Brand Monitoring - ACME Corp",
#     "task_type": 0,
#     "language": "en",
#     "status": 1,
#     "created_at": "2025-01-15T10:30:00.000Z"
#   }
# }
```

--------------------------------

### Implement Simple OR Logic

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Matches posts mentioning either Tesla or BYD.

```json
[{ "in": ["tesla", "byd"] }]
```

--------------------------------

### List Monitoring Tasks with Lenx Open API

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Retrieve a paginated list of monitoring tasks using cURL.

```bash
# List first page of tasks
curl -s "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks?page=1&size=10" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID"
```

--------------------------------

### Paginate Keyword Searches with Python

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Execute any-match keyword searches and retrieve subsequent pages using the search_after parameter.

```python
import os
import requests

endpoint = os.getenv("ADHOCSEARCH_ENDPOINT", "https://adhocsearch.fasta.ai")
headers = {
    "Content-Type": "application/json",
    "x-api-key": os.environ["ADHOCSEARCH_API_KEY"],
    "x-user-id": os.environ["ADHOCSEARCH_USER_ID"],
}

# First page - search for posts containing "election" OR "vote"
resp = requests.post(
    f"{endpoint}/search",
    headers=headers,
    json={
        "keywords": ["election", "vote"],
        "from": 1711929600,
        "to": 1714521600,
        "size": 100,
    },
)
data = resp.json()["data"]
print(f"Found {len(data)} posts")

# Paginate to get more results
if len(data) == 100:
    last_timestamp = data[-1]["unix_timestamp"]
    resp = requests.post(
        f"{endpoint}/search",
        headers=headers,
        json={
            "keywords": ["election", "vote"],
            "from": 1711929600,
            "to": 1714521600,
            "size": 100,
            "search_after": last_timestamp,
        },
    )
    more_data = resp.json()["data"]
    data.extend(more_data)
```

--------------------------------

### Implement Multiple Top-Level Alternatives

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Matches either an EV topic or a solar topic using multiple EQLayer objects.

```json
[
  { "in": [["ev", "electric vehicle"], ["battery"]] },
  { "in": [["solar", "photovoltaic"], ["panel"]] }
]
```

--------------------------------

### Create Task API

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Create a new monitoring task with a search query configuration.

```APIDOC
## POST /api/v1/tasks

### Description
Create a new monitoring task with a search query configuration.

### Method
POST

### Endpoint
{endpoint}/api/v1/tasks

### Parameters
#### Request Body
- **task_type** (string) - Required - "live" or "adhoc"
- **task_name** (string) - Required - 1–50 characters
- **language** (string) - Required - "zh-t", "zh-s", or "en"
- **date_range** (object) - Conditional - Required for `adhoc` tasks, forbidden for `live` tasks
  - **from** (number) - Required - Start timestamp (Unix milliseconds, positive)
  - **to** (number) - Required - End timestamp (Unix milliseconds, must be > `from`)
- **search_query** (object) - Required - Search configuration
  - **query_layer** (EQLayer[]) - Required - At least 1 entry. Each entry has optional `in` and `ex` arrays (see `query_layer` summary below)
  - **region** (string) - Optional - "Hong Kong", "China", "Taiwan", or "USA"
  - **list_medium** (string[]) - Optional - At least 1 when provided. Values: "Facebook", "Instagram", "Social", "News", "Forum", "Blog", "Videos"

### Request Example
```json
{
  "task_type": "adhoc",
  "task_name": "Example Task",
  "language": "en",
  "date_range": {
    "from": 1711929600000,
    "to": 1714521600000
  },
  "search_query": {
    "query_layer": [
      {
        "in": ["term1", "term2"]
      }
    ],
    "region": "USA",
    "list_medium": ["News"]
  }
}
```

### Response
#### Success Response (201 Created)
- **task_id** (number) - The ID of the created task
- **task_name** (string) - The name of the task
- **task_type** (number) - The type of the task (0 for live, 1 for adhoc)
- **language** (string) - The language of the task
- **status** (number) - The status of the task
- **created_at** (string) - The creation timestamp

#### Response Example
```json
{
  "data": {
    "task_id": 123,
    "task_name": "New Task",
    "task_type": 0,
    "language": "en",
    "status": 1,
    "created_at": "2025-01-15T10:30:00.000Z"
  }
}
```

### Errors
- **400**: Invalid request body (validation failed, or `query_layer` could not be converted — `INVALID_QUERY_LAYER`)
- **401**: Missing `x-user-id` header
- **403**: User does not have an associated account
- **404**: User not found
- **422**: Adhoc tasks are not supported yet
- **500**: Unexpected error
```

--------------------------------

### Create Task Response (201 Created)

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Successful response after creating a task.

```json
{ "data": { "task_id": 123, "task_name": "New Task", "task_type": 0, "language": "en", "status": 1, "created_at": "2025-01-15T10:30:00.000Z" } }
```

--------------------------------

### POST /api/v1/tasks

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Create a new live or ad-hoc monitoring task with a search query configuration.

```APIDOC
## POST /api/v1/tasks

### Description
Create a new live or ad-hoc monitoring task with a search query configuration. Live tasks run continuously while ad-hoc tasks require a date range.

### Method
POST

### Endpoint
/api/v1/tasks

### Request Body
- **task_type** (string) - Required - 'live' or 'ad-hoc'.
- **task_name** (string) - Required - Name of the task.
- **language** (string) - Optional - Language code (e.g., 'en').
- **search_query** (object) - Required - Configuration containing `query_layer`, `region`, and `list_medium`.

### Response
#### Success Response (201)
- **task_id** (integer) - The ID of the created task.
- **task_name** (string) - Name of the task.
- **status** (integer) - Current status of the task.
```

--------------------------------

### Implement Lenx Open API Client in TypeScript

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Defines interfaces for post data and configuration, and provides a client class for managing tasks and fetching paginated data. Requires environment variables for endpoint, API key, and user ID.

```typescript
interface LenxPost {
  id: string;
  hash: string;
  medium: string;
  post_message: string;
  post_timestamp: string;
  unix_timestamp: number;
  author_name: string;
  sentiment_score: number | null;
  reaction_count: number | null;
}

interface LenxConfig {
  endpoint: string;
  apiKey: string;
  userId: string;
}

class LenxClient {
  constructor(private config: LenxConfig) {}

  private get headers() {
    return {
      "x-api-key": this.config.apiKey,
      "x-user-id": this.config.userId,
      "Content-Type": "application/json",
    };
  }

  async createTask(name: string, queryLayer: object[], mediums: string[]) {
    const res = await fetch(`${this.config.endpoint}/api/v1/tasks`, {
      method: "POST",
      headers: this.headers,
      body: JSON.stringify({
        task_type: "live",
        task_name: name,
        language: "en",
        search_query: { query_layer: queryLayer, list_medium: mediums },
      }),
    });
    if (!res.ok) throw new Error(`Create failed: ${res.status}`);
    return res.json();
  }

  async getTaskData(taskId: number, from: number, to: number): Promise<LenxPost[]> {
    const allPosts: LenxPost[] = [];
    let searchAfter: number | undefined;

    while (true) {
      const params = new URLSearchParams({
        from: from.toString(),
        to: to.toString(),
        size: "1000",
        ...(searchAfter && { search_after: searchAfter.toString() }),
      });

      const res = await fetch(
        `${this.config.endpoint}/api/v1/tasks/${taskId}/data?${params}`,
        { headers: this.headers }
      );
      if (!res.ok) throw new Error(`Fetch failed: ${res.status}`);

      const { data } = await res.json();
      allPosts.push(...data);

      if (data.length < 1000) break;
      searchAfter = data[data.length - 1].unix_timestamp;
    }
    return allPosts;
  }
}

// Usage
const client = new LenxClient({
  endpoint: process.env.LENX_ENDPOINT ?? "https://open.lenx.ai",
  apiKey: process.env.LENX_API_KEY!,
  userId: process.env.LENX_USER_ID!,
});

const posts = await client.getTaskData(42, 1711929600, 1714521600);
console.log(`Retrieved ${posts.length} posts`);
```

--------------------------------

### List Tasks using curl

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use this `curl` command to list tasks, with support for pagination using `page` and `size` query parameters. Authentication headers and endpoint environment variables are required.

```bash
curl -s "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks?page=1&size=10" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID"
```

--------------------------------

### POST /api/v1/tasks

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Creates a new task with specific configuration.

```APIDOC
## POST /api/v1/tasks

### Description
Creates a new task with the provided configuration, including task type, name, language, and search query parameters.

### Method
POST

### Endpoint
/api/v1/tasks

### Request Body
- **task_type** (string) - Required - The type of task (e.g., "live").
- **task_name** (string) - Required - The name of the task.
- **language** (string) - Required - The language code.
- **search_query** (object) - Required - The search configuration object.

### Request Example
{
  "task_type": "live",
  "task_name": "Brand Monitoring",
  "language": "en",
  "search_query": {
    "query_layer": [
      { "in": [["acme corp", "acme corporation"]], "ex": ["job listing"] }
    ],
    "region": "USA",
    "list_medium": ["News", "Social"]
  }
}
```

--------------------------------

### Implement Include with Exclusions

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Matches Apple content while excluding fruit-related results.

```json
[{ "in": ["apple"], "ex": ["fruit", "recipe"] }]
```

--------------------------------

### Create Task using curl

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use this `curl` command to create a new task. It requires specifying the task type, name, language, and search query parameters. Ensure environment variables for endpoint and authentication are set.

```bash
curl -s -X POST "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID" \
  -H "Content-Type: application/json" \
  -d '{
    "task_type": "live",
    "task_name": "Brand Monitoring",
    "language": "en",
    "search_query": {
      "query_layer": [
        { "in": [["acme corp", "acme corporation"]], "ex": ["job listing"] }
      ],
      "region": "USA",
      "list_medium": ["News", "Social"]
    }
  }'
```

--------------------------------

### Create Task API Endpoint

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use POST to create a new monitoring task. Requires `task_type`, `task_name`, `language`, and `search_query`. `date_range` is conditional based on `task_type`.

```http
POST {endpoint}/api/v1/tasks
```

--------------------------------

### Retrieve Task Data with Pagination

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Fetches monitoring data for a specific task using a loop to handle pagination via the search_after parameter.

```python
import os
import requests

endpoint = os.getenv("LENX_ENDPOINT", "https://open.lenx.ai")
task_id = 42
all_posts = []
search_after = None

while True:
    params = {"from": 1711929600, "to": 1714521600, "size": 1000}
    if search_after is not None:
        params["search_after"] = search_after

    resp = requests.get(
        f"{endpoint}/api/v1/tasks/{task_id}/data",
        headers={
            "x-api-key": os.environ["LENX_API_KEY"],
            "x-user-id": os.environ["LENX_USER_ID"],
        },
        params=params,
    )
    resp.raise_for_status()
    page = resp.json()["data"]
    all_posts.extend(page)

    if len(page) < 1000:
        break
    search_after = page[-1]["unix_timestamp"]

print(f"Retrieved {len(all_posts)} total posts")
# Each post contains: id, hash, country, lang_abbr, medium, channel, post_message,
# post_timestamp, unix_timestamp, author_name, sentiment_score, reaction_count, etc.
```

--------------------------------

### Paginate Task Data using Python

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

This Python script demonstrates how to retrieve task data with pagination. It uses the `requests` library and iterates through results until all data is fetched. Ensure `LENX_ENDPOINT` and `LENX_API_KEY`, `LENX_USER_ID` environment variables are set.

```python
import os, requests

endpoint = os.getenv("LENX_ENDPOINT", "https://open.lenx.ai")
task_id = 42
all_posts, search_after = [], None

while True:
    params = {"from": 1711929600, "to": 1714521600, "size": 1000}
    if search_after is not None:
        params["search_after"] = search_after
    resp = requests.get(
        f"{endpoint}/api/v1/tasks/{task_id}/data",
        headers={"x-api-key": os.environ["LENX_API_KEY"], "x-user-id": os.environ["LENX_USER_ID"]},
        params=params,
    )
    resp.raise_for_status()
    page = resp.json()["data"]
    all_posts.extend(page)
    if len(page) < 1000:
        break
    search_after = page[-1]["unix_timestamp"]
```

--------------------------------

### Implement Grouped AND/OR Logic

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Matches posts mentioning (EV or electric vehicle) AND (battery or charging).

```json
[{ "in": [["ev", "electric vehicle"], ["battery", "charging"]] }]
```

--------------------------------

### Retrieve Task Metadata

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use this endpoint to fetch configuration and metadata for a specific monitoring task.

```http
GET {endpoint}/api/v1/tasks/{task_id}
```

--------------------------------

### Implement Complex Include and Exclude Groups

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Matches AI agents or copilots with automation context, while excluding job-related terms.

```json
[
  {
    "in": [["ai agent", "copilot"], ["automation", "workflow"]],
    "ex": [["job", "jobs", "hiring"]]
  }
]
```

--------------------------------

### Perform a search request with TypeScript

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/fasta-adhocsearch-api/SKILL.md

Uses the fetch API to send a POST request to the AdhocSearch endpoint. Requires environment variables for API key and user ID.

```typescript
const endpoint = process.env.ADHOCSEARCH_ENDPOINT ?? "https://adhocsearch.fasta.ai";
const res = await fetch(`${endpoint}/search`, {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "x-api-key": process.env.ADHOCSEARCH_API_KEY!,
    "x-user-id": process.env.ADHOCSEARCH_USER_ID!,
  },
  body: JSON.stringify({
    query: "artificial intelligence",
    from: 1711929600,
    to: 1714521600,
    size: 25,
  }),
});
const { data } = await res.json();
```

--------------------------------

### List Accessible Tasks

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Retrieve a paginated list of all tasks available to the authenticated user.

```http
GET {endpoint}/api/v1/tasks
```

```json
{
  "data": [
    {
      "task_id": 123,
      "task_name": "My Task",
      "task_type": 0,
      "indus_id": null,
      "labels": [],
      "lang_abbr": "en",
      "status": 1,
      "created_at": "2025-01-15T10:30:00.000Z",
      "updated_at": "2025-03-20T14:00:00.000Z"
    }
  ],
  "pagination": {
    "page": 1,
    "size": 10,
    "total": 42,
    "totalPages": 5
  }
}
```

--------------------------------

### Update Task Response (200 OK)

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Successful response after updating a task.

```json
{ "data": { "task_id": 123, "updated": true } }
```

--------------------------------

### Retrieve Task Information with TypeScript

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Fetch detailed metadata for a specific monitoring task using the Fetch API.

```typescript
const endpoint = process.env.LENX_ENDPOINT ?? "https://open.lenx.ai";
const taskId = 42;

const res = await fetch(`${endpoint}/api/v1/tasks/${taskId}`, {
  headers: {
    "x-api-key": process.env.LENX_API_KEY!,
    "x-user-id": process.env.LENX_USER_ID!,
  },
});

const { data } = await res.json();
// data contains: task_id, task_name, owner_ac_id, task_type, indus_id, labels,
// lang_abbr, status, region, created_at, updated_at, search_query
console.log(`Task: ${data.task_name}, Status: ${data.status}`);
console.log(`Search query:`, data.search_query);
```

--------------------------------

### PATCH Request with query_layer Configuration

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Use this JSON structure to define search parameters including language, media sources, and specific inclusion/exclusion layers. Note that modifying query_layer will automatically trigger a regeneration of the query_string.

```json
{
  "search_query": {
    "lang_abbr": ["en"],
    "list_medium": ["facebook", "news"],
    "query_layer": [
      {
        "in": [["tesla", "byd"], ["battery", "charging"]],
        "ex": ["stock", "share price"]
      }
    ],
    "exclude_channel_links": true
  }
}
```

--------------------------------

### Fetch Task Monitoring Data

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Retrieve monitoring data for a specific task within a defined date range using Unix timestamps.

```http
GET {endpoint}/api/v1/tasks/{task_id}/data
```

--------------------------------

### POST /search

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/fasta-adhocsearch-api/SKILL.md

Executes a search query against the FASTA AdHocSearch service to retrieve social media posts based on specified criteria.

```APIDOC
## POST {endpoint}/search

### Description
Searches social media posts based on a query or keywords within a specified date range and country.

### Method
POST

### Endpoint
{endpoint}/search

### Parameters
#### Request Body
- **from** (number) - Required - Start of date range (Unix timestamp, seconds)
- **to** (number) - Required - End of date range (Unix timestamp, seconds)
- **size** (number) - Required - Number of results to return (1–1000)
- **query** (string) - Optional - Phrase search (max 50 chars). Mutually exclusive with keywords.
- **keywords** (string[]) - Optional - Keyword search, any-match (max 20 items, each max 50 chars). Mutually exclusive with query.
- **search_after** (number) - Optional - Cursor for pagination (unix_timestamp of the last result from the previous page)
- **country** (string) - Optional - Filter by country code

### Request Example
{
  "query": "climate change",
  "from": 1711929600,
  "to": 1714521600,
  "size": 10
}

### Response
#### Success Response (200)
- **data** (array) - List of post objects containing metadata, content, and engagement metrics.

#### Response Example
{
  "data": [
    {
      "hash": "string",
      "post_message": "string",
      "unix_timestamp": 1234567890
    }
  ]
}
```

--------------------------------

### Define Complex Search Logic

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Demonstrates the JSON structure for query layers, supporting OR, AND-of-OR, and exclusion logic.

```json
// Simple OR: Match posts containing "tesla" OR "byd"
[{ "in": ["tesla", "byd"] }]
// Result: (tesla) OR (byd)

// Grouped AND/OR: Match (EV or electric vehicle) AND (battery or charging)
[{ "in": [["ev", "electric vehicle"], ["battery", "charging"]] }]
// Result: ((ev) OR (electric vehicle)) AND ((battery) OR (charging))

// Include with exclusions: Apple content excluding fruit results
[{ "in": ["apple"], "ex": ["fruit", "recipe", "pie"] }]
// Result: ((apple)) AND NOT ((fruit) OR (recipe) OR (pie))

// Multiple alternatives: Match EV topic OR solar topic
[
  { "in": [["ev", "electric vehicle"], ["battery"]] },
  { "in": [["solar", "photovoltaic"], ["panel"]] }
]
// Result: (EV AND battery) OR (solar AND panel)

// Complex grouped query: AI agents with automation, excluding job ads
[
  {
    "in": [["ai agent", "copilot"], ["automation", "workflow"]],
    "ex": [["job", "jobs", "hiring"]]
  }
]
// Result: ((ai agent) OR (copilot)) AND ((automation) OR (workflow)) AND NOT ((job) OR (jobs) OR (hiring))
```

--------------------------------

### Delete Task Response (200 OK)

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Successful response after deleting a task.

```json
{ "data": { "task_id": 123, "deleted": true } }
```

--------------------------------

### API Endpoint Definition

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/fasta-adhocsearch-api/SKILL.md

The base endpoint for performing search requests.

```text
POST {endpoint}/search
```

--------------------------------

### Update Task Configuration

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Modifies an existing task's name or search query parameters using a PATCH request.

```bash
# Update task search query to monitor EV industry
curl -s -X PATCH "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks/42" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID" \
  -H "Content-Type: application/json" \
  -d '{
    "task_name": "EV Industry Monitor",
    "search_query": {
      "lang_abbr": ["en"],
      "list_medium": ["facebook", "news", "forum"],
      "query_layer": [
        {
          "in": [["tesla", "byd", "rivian"], ["battery", "charging", "ev"]],
          "ex": ["stock price", "share price", "investor"]
        }
      ],
      "exclude_channel_links": true
    }
  }'

# Response (200 OK):
# { "data": { "task_id": 42, "updated": true } }
```

--------------------------------

### Search Social Media Posts with FASTA AdHocSearch

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Perform an ad-hoc search for posts using a query string and date range via cURL.

```bash
# Search for posts containing "climate change" in the last 30 days
curl -X POST "${ADHOCSEARCH_ENDPOINT:-https://adhocsearch.fasta.ai}/search" \
  -H "Content-Type: application/json" \
  -H "x-api-key: $ADHOCSEARCH_API_KEY" \
  -H "x-user-id: $ADHOCSEARCH_USER_ID" \
  -d '{
    "query": "climate change",
    "from": 1711929600,
    "to": 1714521600,
    "size": 10
  }'
```

--------------------------------

### PATCH /api/v1/tasks/{task_id}

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Update a task's name or search configuration.

```APIDOC
## PATCH /api/v1/tasks/{task_id}

### Description
Update a task's name or search configuration. The `query_layer` provides structured search logic.

### Method
PATCH

### Endpoint
/api/v1/tasks/{task_id}

### Parameters
#### Path Parameters
- **task_id** (integer) - Required - The ID of the task to update.

### Request Body
- **task_name** (string) - Optional - New name for the task.
- **search_query** (object) - Optional - Updated search configuration.
```

--------------------------------

### POST /search

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Search for social media posts using query phrases or keywords within a specific date range.

```APIDOC
## POST /search

### Description
Enables searching social media posts by query phrases or keywords within a specified date range. Supports filtering by country and returns post data including engagement metrics.

### Method
POST

### Endpoint
/search

### Parameters
#### Request Body
- **query** (string) - Optional - The search phrase.
- **keywords** (array) - Optional - List of keywords for any-match search.
- **from** (integer) - Required - Start timestamp.
- **to** (integer) - Required - End timestamp.
- **size** (integer) - Required - Number of results to return.
- **search_after** (integer) - Optional - Unix timestamp of the last result for pagination.

### Request Example
{
  "query": "climate change",
  "from": 1711929600,
  "to": 1714521600,
  "size": 10
}

### Response
#### Success Response (200)
- **data** (array) - List of post objects containing hash, country, lang_abbr, medium, channel, post_message, post_timestamp, unix_timestamp, author_name, reaction_count, share_count, and comment_count.
```

--------------------------------

### Update Task API

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Update an existing task with partial updates.

```APIDOC
## PATCH /api/v1/tasks/{task_id}

### Description
Update an existing task with partial updates. At least one of `task_name` or `search_query` must be provided.

### Method
PATCH

### Endpoint
{endpoint}/api/v1/tasks/{task_id}

### Parameters
#### Path Parameters
- **task_id** (number) - Required - The ID of the task to update

#### Request Body
- **task_name** (string) - Optional - 1–50 characters
- **search_query** (object) - Optional - Search query configuration
  - **lang_abbr** (string | null) - Optional
  - **region** (string | null) - Optional
  - **list_medium** (string[]) - Optional
  - **list_author_id** (string[]) - Optional
  - **query_layer** (EQLayer[] | null) - Optional - Structured search logic. Pass `null` to clear.

> **Validation:** At least one of `task_name` or `search_query` must be provided. If `search_query` is provided, it must contain at least one field. Sending an empty body is invalid.

### Request Example
```json
{
  "task_name": "Updated Task Name",
  "search_query": {
    "list_medium": ["News", "Blog"]
  }
}
```

### Response
#### Success Response (200 OK)
- **task_id** (number) - The ID of the updated task
- **updated** (boolean) - Indicates if the task was updated

#### Response Example
```json
{
  "data": {
    "task_id": 123,
    "updated": true
  }
}
```

### Errors
- **400**: Invalid request body (validation failed, or `query_layer` could not be converted — `INVALID_QUERY_LAYER`)
- **401**: Missing `x-user-id` header
- **403**: User does not have permission to access this task
- **404**: Task not found or Search query not found for this task
- **500**: Unexpected error
```

--------------------------------

### Delete Monitoring Task

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Performs a soft-delete on a task, restricted to the task owner.

```bash
# Delete a monitoring task
curl -s -X DELETE "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks/42" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID"

# Response (200 OK):
# { "data": { "task_id": 42, "deleted": true } }

# Error responses:
# 403: User is not the task owner
```

--------------------------------

### PATCH /api/v1/tasks/{task_id}

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/references/QUERY_LAYER.md

Updates a monitoring task by providing a new query_layer structure. The backend will regenerate the internal query_string based on this input.

```APIDOC
## PATCH /api/v1/tasks/{task_id}

### Description
Updates the search logic for a specific monitoring task using the query_layer structure.

### Method
PATCH

### Endpoint
/api/v1/tasks/{task_id}

### Parameters
#### Path Parameters
- **task_id** (string) - Required - The unique identifier of the task to update.

#### Request Body
- **query_layer** (array) - Required - An array of EQLayer objects defining the search logic. Each EQLayer contains 'in' (terms to include) and 'ex' (terms to exclude) fields, which accept either string[] or string[][] types.

### Request Example
{
  "query_layer": [
    {
      "in": [["ai agent", "copilot"], ["automation", "workflow"]],
      "ex": [["job", "jobs", "hiring"]]
    }
  ]
}

### Response
#### Success Response (200)
- **status** (string) - Confirmation of the update.

#### Response Example
{
  "status": "success"
}
```

--------------------------------

### Update Task API Endpoint

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use PATCH to partially update an existing task. All fields in the body are optional, but at least one of `task_name` or `search_query` must be provided.

```http
PATCH {endpoint}/api/v1/tasks/{task_id}
```

--------------------------------

### Update Task Name using curl

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use this `curl` command to update the name of an existing task. Ensure `LENX_ENDPOINT` and `LENX_API_KEY`, `LENX_USER_ID` environment variables are set.

```bash
curl -s -X PATCH "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks/42" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID" \
  -H "Content-Type: application/json" \
  -d '{"task_name": "New Task Name"}'
```

--------------------------------

### API Response Schema

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/fasta-adhocsearch-api/SKILL.md

The structure of the JSON response returned by the API.

```json
{
  "data": [
    {
      "hash": "string",
      "country": "string",
      "lang_abbr": "string",
      "medium": "string",
      "channel": "string",
      "channel_link": "string",
      "site": "string",
      "thread_link": "string",
      "post_link": "string",
      "thread_title": "string",
      "post_message": "string",
      "post_timestamp": "ISO 8601 datetime",
      "unix_timestamp": 1234567890,
      "author_name": "string",
      "author_id": "string | null",
      "author_image": "string | null",
      "author_link": "string | null",
      "is_comment": false,
      "comment_order": "number | null",
      "comment_count": "number | null",
      "share_count": "number | null",
      "view_count": "number | null",
      "reaction_count": "number | null",
      "reaction_like": "number | null",
      "reaction_dislike": "number | null",
      "reaction_love": "number | null",
      "reaction_wow": "number | null",
      "reaction_haha": "number | null",
      "reaction_angry": "number | null"
    }
  ]
}
```

--------------------------------

### DELETE /api/v1/tasks/{task_id}

Source: https://context7.com/thinkcol/fasta-skills/llms.txt

Soft-delete a monitoring task.

```APIDOC
## DELETE /api/v1/tasks/{task_id}

### Description
Soft-delete a monitoring task. Only the task owner can perform deletion.

### Method
DELETE

### Endpoint
/api/v1/tasks/{task_id}

### Parameters
#### Path Parameters
- **task_id** (integer) - Required - The ID of the task to delete.

### Response
#### Success Response (200)
- **task_id** (integer) - The ID of the deleted task.
- **deleted** (boolean) - Confirmation of deletion.
```

--------------------------------

### Delete Task using curl

Source: https://github.com/thinkcol/fasta-skills/blob/master/skills/open-lenx-api/SKILL.md

Use this `curl` command to perform a soft delete on a task. This operation requires the task ID and authentication headers. Ensure environment variables for endpoint and authentication are set.

```bash
curl -s -X DELETE "${LENX_ENDPOINT:-https://open.lenx.ai}/api/v1/tasks/42" \
  -H "x-api-key: $LENX_API_KEY" \
  -H "x-user-id: $LENX_USER_ID"
```