Catches the mistakes coding agents repeat before they merge: a pull-request gate with no model call and no API key.
OpenClaw loads skills from the workspace’s skills folder, so brevio-core reaches it as a skill, and as an MCP server
if you add one.
In the repository:
brevio-core hook install --target openclaw
brevio-core init
hook install writes the agent skill to skills/brevio-core/SKILL.md. Its front matter asks for the brevio-core
command, so OpenClaw shows the skill only where the package is installed. init writes the pull-request workflow, the
AGENTS.md block and a baseline.
The same skill is published to the skill registry with each release, so it can be installed from there too.
Add brevio-core where your OpenClaw configuration lists MCP servers, with this command and these arguments:
{ "brevio-core": { "command": "uvx", "args": ["brevio-core==0.9.1", "mcp"] } }
scan, the agent hook and the MCP server use no network and send nothing.
The skill tells the agent to scan the files it changed with brevio-core scan and to fix each finding in the code
rather than silence it. There is no hook after each edit, so the pull-request check is the backstop.