git:20260717.2cecabb to git:20260717.3220402

4 added, 13 removed. Audit A to A.

browser-harness is a thin layer that connects agents to browsers via an editable CDP harness.

# Code priorities

- Clarity

- Precision

- Low verbosity

- Versatility

# Overview

Core code lives in `src/browser_harness/`:

- `admin.py` — daemon lifecycle, diagnostics, updates, profile management

- `daemon.py` — the long-lived middleman process between the browser and the agent

- `helpers.py` — CDP wrapper and core browser primitives auto-imported into `-c` scripts

- `run.py` — the `browser-harness` CLI

`SKILL.md` tells agents how to use the harness and CLI.

`install.md` tells agents how to install it, attach a browser, and troubleshoot.

In this checkout, invoke the current source with `./browser-harness`; do not use

a globally installed `browser-harness` binary.

- For any session-recording or video task, read

- `skills/browser-harness-video/SKILL.md` fully and reuse the shared renderer.

- Do not invent a separate editing workflow.

- Recording and video creation require user intent. Treat any natural-language

- nudge such as “show me what you did,” “record this,” “make a video,” “demo it,”

- or “walk me through it” as opt-in; do not require exact wording. Significant

- work alone is not consent. For a video requested before browser work, call

- `start_recording()` before the first action and retain its exact returned path;

- never substitute `./browser-harness recordings --latest`. Use `--latest` only

- for a post-task request after verifying timestamps and pages. Never reenact a

- task. When a video was requested and subagents are available, delegate only

- post-production with the original task and exact recording path.

+ For recording or video tasks, follow `SKILL.md` and

+ `interaction-skills/make-video.md`. A natural request to show, record, or demo

+ the work opts in; significant work alone does not. Keep the exact path returned

+ by `start_recording()` and never reenact a finished task.

An agent operating the harness only edits inside `agent-workspace/`:

- `agent_helpers.py` — task-specific browser helpers the agent adds

- `domain-skills/` — skills the agent writes and reads

# Contributing

Consider what is really needed. Prefer the smallest diff that fixes the bug.