watchlight-claude-agent

Governed Claude Agent SDK agents with Watchlight — authorize every agent action, fail-closed, with zero infrastructure.

pip install watchlight-claude-agent

Independent third-party plugin. This is an independent integration built by Watchlight AI. It is not affiliated with, endorsed by, or sponsored by Anthropic. Claude Agent SDK and related names are trademarks of Anthropic, used here nominatively only to describe compatibility.

What it does

watchlight-claude-agent puts a Watchlight authorization decision in front of every action your Claude Agent SDK agent takes — tool calls, plan steps, and native sub-agent spawns — so each one is allowed, modified, or denied before it runs. It's open-source glue: a thin, framework-specific layer that threads Watchlight's governance primitives into your agent's hook loop. The actual policy decisions run on Watchlight's compiled engine, either in-process for local development or against the governed control plane in production.

Quickstart

Point the plugin at a backend and wrap your run. Your agent code doesn't change — only the backend does.

For local development, the zero-infrastructure Developer Edition runs the compiled engine in-process (requires the watchlight-engine package):

pip install watchlight-claude-agent watchlight-engine

from watchlight_claude_agent import WatchlightClaudeAgentSDKPlugin
from watchlight_core import InProcessClient

# A Cedar policy: the research agent may read, nothing else.
POLICIES = [
    {"name": "reader",
     "code": 'permit(principal == User::"research-agent", action == Action::"read", resource);'},
]

plugin = WatchlightClaudeAgentSDKPlugin()
plugin.apdp = InProcessClient(POLICIES)   # decisions run in-process, no server, no network

async def run_agent(question: str):
    async with await plugin.start_run("research-agent") as handle:
        # Validate a multi-step plan ahead of execution (Allow / Modify / Deny).
        result = await handle.submit_plan(["read dataset", "summarize"])
        if result.is_deny():
            raise PermissionError(f"Plan denied: {result.violations}")

        # Gate each action authoritatively. Returns True on Allow, False on Deny.
        if not await handle.authorize_action("read", "dataset"):
            raise PermissionError("Denied by policy")

        # ... your Claude agent runs, every action governed ...

authorize_action fails closed: a denial — or an unreachable backend — returns False, so the action never runs.

Two backends, same code

Moving from local to production is a one-line change:

from watchlight_core import ApdpClient

plugin = WatchlightClaudeAgentSDKPlugin()
plugin.apdp = ApdpClient("https://apdp.your-company.example", api_key="...")

Links

• Documentation: https://docs.watchlight.ai/de
• Website: https://watchlight.ai

License

Apache-2.0