0
Run ci
Tüm testler başarılı olana kadar CI kontrollerini çalıştırın ve hataları düzeltin
yazan kuscubasiece·8 Eylül 2026
Kaynak: davepoon/claude-code-subagents-collection · MIT
Tüm testler başarılı olana kadar CI kontrollerini çalıştırın ve hataları düzeltin
Kaynak: davepoon/claude-code-subagents-collection · MIT
mkdir -p .claude/commands && curl -sL https://www.antrophiclove.com/api/kaynak/run-ci/raw -o .claude/commands/run-ci.mdHedef dosya: .claude/commands/run-ci.md
---
description: Run CI checks and fix any errors until all tests pass
category: ci-deployment
allowed-tools: Bash, Edit, Read, Glob
---
Run CI checks for the project and fix any errors until all tests pass.
## Process:
1. **Detect CI System**:
- Check for CI configuration files:
- `.github/workflows/*.yml` (GitHub Actions)
- `.gitlab-ci.yml` (GitLab CI)
- `.circleci/config.yml` (CircleCI)
- `Jenkinsfile` (Jenkins)
- `.travis.yml` (Travis CI)
- `bitbucket-pipelines.yml` (Bitbucket)
2. **Detect Build System**:
- JavaScript/TypeScript: package.json scripts
- Python: Makefile, tox.ini, setup.py, pyproject.toml
- Go: Makefile, go.mod
- Rust: Cargo.toml
- Java: pom.xml, build.gradle
- Other: Look for common CI scripts
3. **Run CI Commands**:
- Check for CI scripts: `ci`, `test`, `check`, `validate`, `verify`
- Common script locations:
- `./scripts/ci.sh`, `./ci.sh`, `./run-tests.sh`
- Package manager scripts (npm/yarn/pnpm run test)
- Make targets (make test, make ci)
- Activate virtual environments if needed (Python, Ruby, etc.)
4. **Fix Errors**:
- Analyze error output
- Fix code issues, test failures, or configuration problems
- Re-run CI checks after each fix
5. **Common CI Tasks**:
- Linting/formatting
- Type checking
- Unit tests
- Integration tests
- Build verification
- Documentation generation
## Examples:
- JavaScript: `npm test` or `npm run ci`
- Python: `make test` or `pytest` or `tox`
- Go: `go test ./...` or `make test`
- Rust: `cargo test`
- Generic: `./ci.sh` or `make ci`
Continue fixing issues and re-running until all CI checks pass.