Replace Deployment Checklists
With A Single Workflow
Drop one Docker image into your pipeline. Get AI code review, deployment approvals, and Slack notifications — without new infrastructure or dedicated agents.
Why teams choose Shiro
Shiro is not another CI platform — it's the AI orchestration layer that runs inside your existing infrastructure. No SaaS, no lock-in, no orchestration clusters.
Turn your CI runner into an AI workflow engine
Runs as a Docker image inside GitLab CI or GitHub Actions. No new servers, no agents, no infra to manage.
AI reviews, approvals, and deployments running in your existing CI/CD
AI code review, release notes, incident summaries — powered by OpenAI, Ollama, or any custom endpoint.
Need manager approval before production deploy? Add one step
Pause a workflow mid-pipeline, send a Slack notification, and resume on manual GitLab job trigger.
Single Binary
One 10MB Go binary. No JVM, no runtime dependencies. Works inside any CI runner that has Docker.
Module System
Install Slack, Jira, git, AI modules with one command. Community modules on GitHub.
State Persistence
Save workflow state to GitLab artifacts between pipeline stages. Resume exactly where you left off.
Built for modern infrastructure teams
Shiro is designed for teams who need AI automation without the operational overhead of dedicated orchestration infrastructure.
Platform Engineers
Build internal developer platforms with AI workflows that run in your existing CI.
DevOps Teams
Stop maintaining deployment checklists. Automate everything with workflows.
AI Engineering
Ship AI-powered workflows with governance and human-in-loop controls.
Enterprise Teams
Production deployment governance without another platform.
Not just another CI tool
Shiro fills the gap between CI platforms and AI orchestration — without the complexity of dedicated infrastructure.
vs Jenkins/GitLab CI
- ✓AI-native workflows built-in
- ✓Human-in-loop approvals with state persistence
- ✓Portable across CI platforms
- ✓No pipeline YAML complexity
vs Temporal/n8n
- ✓Runs inside your existing CI (no new servers)
- ✓Single binary, no orchestration cluster
- ✓Ephemeral execution (no long-running state machines)
- ✓Zero operational overhead
vs LangChain
- ✓Built for CI/CD workflows, not just AI chains
- ✓Enterprise governance and approvals
- ✓Native GitLab/GitHub integration
- ✓Production-ready error handling
In your pipeline in minutes
No install step. No binary download. Use the Docker image directly in your GitLab CI or GitHub Actions job.
- ✓Works in any GitLab CI runner
- ✓Works in any GitHub Actions runner
- ✓State saved to artifacts between stages
- ✓Full docs at shiro-docs.rajit.cc
ai-review:
image: ghcr.io/rajitk13/shiro-automation:latest
variables:
GITLAB_TOKEN: $GL_TOKEN
script:
- shiro run -workflow .shiro/workflows/code-review.json \
-config .shiro/config.yaml -state-store gitlab
only:
- merge_requestsai-review:
runs-on: ubuntu-latest
container:
image: ghcr.io/rajitk13/shiro-automation:latest
steps:
- uses: actions/checkout@v4
- run: shiro run -workflow .shiro/workflows/code-review.jsonWhat you can automate
Pre-built workflow examples for the most common CI automation needs.
AI PR / MR Review
Posts an AI-generated code review comment on every merge request.
Deployment Approval
Slack notification + human gate before production deploys.
Push Notifications
Notify your team on every push to main via Slack.
Release Notes
Auto-generate release notes from commits using AI.
Jira Automation
Create or update Jira issues from CI events automatically.
Incident Summaries
Summarize failures and alert the right team instantly.
Add AI to your CI today
One Docker image. Works in GitLab CI and GitHub Actions. Open source, Apache 2.0.