From business-agent
Use this skill when the user is working in or asking about the business-agent/ directory, editing business-agent.agent.yaml or business-agent.environment.yaml, running managed-agent sessions, or debugging Claude Agent SDK session/vault/resource setup.
How this skill is triggered — by the user, by Claude, or both
Slash command
/business-agent:business-agentThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
The business-agent is a Claude-managed agent that scans GitHub repositories for open issues
The business-agent is a Claude-managed agent that scans GitHub repositories for open issues and opens fix PRs. It runs via the Claude Agent SDK.
| File | Purpose |
|---|---|
business-agent.agent.yaml | Agent definition (model, tools, MCP servers, skills) |
business-agent.environment.yaml | Runtime environment (compute, mounts, timeouts) |
setup.sh | One-time setup — creates agent + environment on the platform, writes IDs to .env |
run_agent.py | Triggers a full run against the created agent/environment |
requirements.txt | Python deps (anthropic SDK) |
.env | Not committed — holds AGENT_ID and ENV_ID written by setup.sh |
run_agent.py| Var | Source |
|---|---|
AGENT_ID | Written to .env by setup.sh |
ENV_ID | Written to .env by setup.sh |
GITHUB_TOKEN | Personal access token (Contents read+write) for repo cloning and push |
GITHUB_MCP_TOKEN | OAuth token for the GitHub Copilot MCP endpoint (scopes: repo, issues, pull_requests) |
After editing business-agent.agent.yaml, deploy the new version with:
source .env
ant beta:agents update --agent-id "$AGENT_ID" < business-agent.agent.yaml
Do not re-run setup.sh — that creates a new agent/environment and overwrites .env.
sjbrenchley89/claude-codesjbrenchley89/source-build-ausjbrenchley89/windows-mcpsjbrenchley89/ruflosjbrenchley89/tailscaleAfter each run, outputs (including digest.md) are written to /mnt/session/outputs/ and
downloaded by run_agent.py to the local working directory.
npx claudepluginhub sjbrenchley89/claude-code --plugin business-agentGuides collaborative design exploration before implementation: explores context, asks clarifying questions, proposes approaches, and writes a design doc for user approval.
Creates structured, bite-sized implementation plans from specs or requirements before writing code. Useful for breaking down multi-step tasks into testable steps with file structure and task boundaries.
Implements work from a spec or tickets using TDD at agreed seams, with regular typechecking and test runs, followed by code review.