tama_http
HTTP tools let agents fetch URLs and post data to APIs. They must be unlocked via a skill.
tama_http_get
Section titled “tama_http_get”Fetches the content of a URL via HTTP GET.
Parameters
| Name | Type | Description |
|---|---|---|
url | string | The URL to fetch |
Returns the full response body as a string.
Example skill
---name: fetch-urldescription: Fetch the content of any URL.tools: [tama_http_get]---
Call tama_http_get with the full URL including query parameters.The response body is returned as a string. For HTML pages, extractrelevant text by looking for content inside `<p>`, `<article>`, or`<main>` tags.tama_http_post
Section titled “tama_http_post”Sends an HTTP POST request with a JSON body.
Parameters
| Name | Type | Description |
|---|---|---|
url | string | The URL to post to |
body | string | JSON body as a string |
Returns the full response body as a string.
Example skill
---name: call-apidescription: Call a JSON API endpoint with a POST request.tools: [tama_http_post]---
Use tama_http_post(url, body) where body is a JSON string.Example: tama_http_post("https://api.example.com/data", '{"key": "value"}')Common pattern: web search
Section titled “Common pattern: web search”The canonical search-web skill combines both HTTP tools:
---name: search-webdescription: Search the web using DuckDuckGo. No API key required.tools: [tama_http_get, tama_files_write]---
Call tama_http_get with this URL pattern: https://html.duckduckgo.com/html/?q=<url-encoded-query>
Encode spaces as `+`. Extract results from `<a class="result__a">` and`<span class="result__snippet">`. Follow interesting URLs with anothertama_http_get call to read the full page.