Load agent skills and code-mode guidance

The agentbrowser.agent module keeps named browser controllers across calls in one Python process. The wheel carries an Agent Plugin with Python-specific browser instructions and the pinned engine's native skills.

Keep a controller across calls

import agentbrowser.agent as browser_agent

print(browser_agent.help())

browser_agent.help() returns a short SDK workflow and the paths to the Agent Plugin resources installed with the same pyagentbrowser version.

Create a named module-owned controller when browser work spans code-mode calls:

browser = browser_agent.create("research")
browser.page.open("https://example.com")
page = browser.page.observe()
print(page.origin)
print(page.text)

# In a later code-mode call
browser = browser_agent.get("research")

# After the final call
browser_agent.close("research")

Use browser_agent.names() to inspect registered controller names. Call browser_agent.close_all() after a task that created several controllers.

Pass session=agentbrowser.SessionOptions(...) to create() when the task needs domain restrictions, confirmation policy, or another session option. Duplicate names raise ValueError. get() retrieves one open controller and close() releases it.

Read Code-mode integration to use browser objects and screenshot content in an existing execution host.

Load the Python instructions

plugin = browser_agent.agent_plugin()
skill = browser_agent.agent_skill()

print(plugin)
print(skill.file("SKILL.md"))
instructions = skill.body

The packaged pyagentbrowser skill uses Browser, AsyncBrowser, typed namespaces, and browser.native directly. These Python operations share the embedded native action service with the upstream engine.

Discover the capability in Marimo

Marimo, a Python notebook environment, discovers installed code-mode capabilities through package entry points in version 0.24.0 or later. Install pyagentbrowser in the notebook environment, then inspect its entry:

import marimo._code_mode as cm

print(cm.capabilities()["pyagentbrowser"])

Capability discovery exposes the Python module and its instructions. The code-execution host supplies the application URL and forwards screenshot content through its image channel.

Read the embedded native instruction files

An agent skill is a Markdown instruction set with optional supporting files. The embedded native skills let an application supply version-matched upstream guidance to an agent.

import agentbrowser.skills as skills

print(skills.available())
print(skills.read("core"))

The core skill teaches the engine's browser interaction workflow. read(name) returns its SKILL.md content.

Inspect supporting parts

Function	Result
available(*, include_hidden=False)	Skill names
list(*, full=False, include_hidden=False)	Skill metadata, optionally with loaded files
get(name, *, full=False)	One Skill
parts(name)	SkillPart metadata for every file
part(name, path)	One SkillFile
read(name, path="SKILL.md")	One file's text
markdown(name, *, full=False)	Main instructions or combined skill Markdown

Use the main file when the agent needs the normal browser workflow. Load individual parts when a task reaches the capability they describe. Use full=True when the host needs to package the complete instruction tree in one operation.

Skill paths reject absolute paths, traversal, empty components, and malformed relative paths. Missing skills or files raise KeyError.

Skill records its name, frontmatter description, main content, part metadata, loaded files, and hidden state. SkillPart records a relative path and kind. SkillFile records a relative path and text content.