An AI Agent that quickly solves simple GitHub issues
Powered by Large Language Models (LLMs).
When a developer creates an issue in a repository and passes to this agent, it autonomously works to solve the issue and submits its results as a Pull Request on GitHub.
- Data transmission is limited to the LLM provider and GitHub only
- Only execution of predefined tools is allowed, not arbitrary shell execution
- Creates a new working branch from the base branch for each task and submits pull requests - cannot push to or merge into the base branch
issue-agent help$ issue-agent create-pr clover0/example-repository/issues/123 \
--base_branch main \
--model claude-sonnet-4-20250514- Set up the workflow file in your repository.
name: Run Agent
on:
issues:
# This workflow is triggered when an issue is labeled with 'run-agent'
types:
- labeled
create-pr:
if: ${{ github.event.label.name == 'run-agent' }}
name: Create PR
runs-on: ubuntu-latest
steps:
- name: Install Issue Agent
uses: clover0/setup-issue-agent@v1
- name: configure aws credentials
uses: aws-actions/configure-aws-credentials@v4
with:
role-to-assume: "arn:aws:iam::<AWS-ACCOUNT>:role/<ROLE-NAME>"
role-session-name: run-agent-${{ github.run_id }}
aws-region: "<AWS-REGION>"
- uses: actions/create-github-app-token@v2
id: app-token
with:
app-id: ${{ secrets.TOKEN_APP_ID }}
private-key: ${{ secrets.TOKEN_APP_PRIVATE_KEY }}
- name: Run Issue Agent Action
run: |
issue-agent create-pr ${GITHUB_REPOSITORY}/issues/${{ github.event.issue.number }} \
--base_branch main \
--model us.anthropic.claude-sonnet-4-20250514-v1:0 \
--aws_region us-east-1
env:
GITHUB_TOKEN: ${{ steps.app-token.outputs.token }}See setup-issue-agent for more details.
-
Label the issue with
run-agentto trigger the workflow. -
Issue Agent will create a Pull Request.
Refer to the documentation for more details.
-
Fully Autonomous
- Handles simple coding and documentation tasks without human intervention.
-
Minimal Configuration
- Easy to set up: runs locally or via GitHub Actions in minutes.
-
Security
- Limited scope: cannot execute arbitrary or unsafe code (only predefined and controlled functions).
- The agent running in a container is only provided with the necessary environment variables, which minimizes the risk of credential leakage
Issue Agent makes life easier, especially for routine or repetitive tasks:
- Routine Development Tasks (e.g., basic migration, formatting, simple refactoring tasks)
- Documentation Maintenance (generate or update consistently formatted docs)
- Code Cleanup (e.g., batch updates, typo corrections, simple bug fixes)
We recommend Anthropic's Claude models for optimal performance:
| Provider | Supported Models |
|---|---|
| OpenAI | gpt-4o, gpt-4o-mini |
| Anthropic⭐️ | claude-sonnet-4-20250514, claude-opus-4-20250514, claude-3-5-sonnet-latest, claude-3-7-sonnet-20250219 |
| AWS Bedrock⭐️ | See AWS Bedrock section |
The following models are supported.
- claude-sonnet-4 (ModelID = us.anthropic.claude-sonnet-4-20250514-v1:0, Cross-region inference)
- claude-opus-4 (ModelID = us.anthropic.claude-opus-4-20250514-v1:0, Cross-region inference)
- claude-3-7-sonnet (ModelID = anthropic.claude-3-7-sonnet-20250219-v1:0)
- claude-3-7-sonnet (ModelID = us.anthropic.claude-3-7-sonnet-20250219-v1:0, Cross-region inference)
- claude-3-5-sonnet v2 (ModelID = anthropic.claude-3-5-sonnet-20241022-v2:0)
- claude-3-5-sonnet v2 (ModelID = us.anthropic.claude-3-5-sonnet-20241022-v2:0, Cross-region inference)
- claude-3-5-sonnet v1 (ModelID = anthropic.claude-3-5-sonnet-20240620-v1:0)
- claude-3-5-sonnet v1 (ModelID = us.anthropic.claude-3-5-sonnet-20240620-v1:0, Cross-region inference)