Triage issues and open fix PRs with a GitHub bot
Build a Muse Spark agent that acts on GitHub issues and pull requests. Task: Suggested implementation: Build a GitHub bot using Muse Spark and OpenCode to autonomously triage issues, review PRs, and answer questions. For code fixes, the agent only acts when a maintainer triggers it via a specific comment or label, then writes a patch and test, verifies it, and opens a PR. Human maintainers retain full control over merging and approving changes.; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: Requires a GitHub repository with secrets and workflows, a Meta Model API key, and OpenCode setup. The agent is tuned for Python/pytest/uv stacks and requires adaptation for other languages. Security constraints include no network egress and strict bash allow-lists. Human approval is mandatory for all code merges..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
Build a Muse Spark agent that acts on GitHub issues and pull requests.
Workflow
Suggested implementation: Build a GitHub bot using Muse Spark and OpenCode to autonomously triage issues, review PRs, and answer questions. For code fixes, the agent only acts when a maintainer triggers it via a specific comment or label, then writes a patch and test, verifies it, and opens a PR. Human maintainers retain full control over merging and approving changes.
Reported result
The source describes the intended workflow and design principles but does not report measured outcomes, success rates, or specific business impact from actual deployments. No quantitative results are available to confirm performance or failure rates.
Human role
The original post does not say how a person stayed involved.
Limitations
Requires a GitHub repository with secrets and workflows, a Meta Model API key, and OpenCode setup. The agent is tuned for Python/pytest/uv stacks and requires adaptation for other languages. Security constraints include no network egress and strict bash allow-lists. Human approval is mandatory for all code merges.
Try it
Act as a DevOps engineer setting up a GitHub automation bot. I need to implement a Muse Spark agent using OpenCode to handle GitHub issues and PRs for my Python repository. 1. **Setup**: Create the necessary configuration files (`opencode.json`, `AGENTS.md`, and `.opencode/agent/*.md`) in my repo root. Ensure `AGENTS.md` includes the security section preventing self-modification and treating issue text as untrusted data. 2. **Agents**: Define agents for `triage` (labeling issues), `review` (checking PRs for style and AI slop), and `bugfix` (writing patches only when triggered by a maintainer's `/oc` comment or `agent-fix` label). 3. **Constraints**: Configure the `bugfix` agent to open a PR but never merge it. Ensure `webfetch` is disabled and `bash` commands are restricted to an allow-list (e.g., `git`, `pytest`, `gh pr create`). 4. **Verification**: Before deploying to production, run a dry test on a sample issue and a sample PR. Show me the generated comments and the proposed PR diff. 5. **Human Gate**: Do not apply any changes to the main branch or merge any PRs. Wait for my explicit confirmation after reviewing the dry run results. If any configuration files are missing or if the stack is not Python, ask me for clarification before proceeding.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.