-
-
logo
About the Project Inspiration As a founder and AI engineer, I watched development teams struggle with the same bottleneck over and over. Developers were merging code faster than humans could review it. Security vulnerabilities slipped through. Quality issues piled up. The problem wasn't the developers......... it was the process. There was no scalable way to give every merge request the attention it deserved. Sprint Guardian was born from that gap.
What if every developer had a world class security reviewer, quality engineer, and technical lead available on every single MR, automatically?
What It Does Sprint Guardian is a four agent merge request review pipeline that automatically deploys four specialized AI agents in sequence when a developer opens a merge request:
🔍 Triage Agent — reads the full diff, assigns risk level, creates a problem marker every downstream agent works from 🔒 Security Gate Agent — scans for hardcoded secrets, auth flaws, injection risks, and insecure config. Rates each PASS, WARN, or FAIL ✅ Quality Check Agent — reviews correctness, readability, testability, and maintainability 🔓 Unblock Agent — synthesizes all three upstream reports into one prioritized action list for the developer
The result is a complete security and quality report posted directly on the MR — telling the developer exactly what to fix and where.
How We Built It Sprint Guardian is built using a multi agent orchestration architecture with four specialized AI agents powered by Anthropic Claude, coordinated through a sequential flow pipeline. Each agent has a dedicated SKILL.md file that defines:
Its identity and purpose Step by step job instructions Risk criteria and severity ratings Explicit guardrails A self check criteria it must pass before posting output
This agent identity framework is derived from our proprietary Staged Intelligence methodology, a four phase multi tier AI orchestration system developed and trade secret protected by MKE BIZ TECH LLC.
Challenges We Ran Into The biggest challenge was schema debugging. Early versions of the flow were written in incorrect syntax, meaning the pipeline sat silently doing nothing. Debugging from scratch, understanding the correct agent orchestration structure, and getting skill files wired correctly took significant iteration. Getting agents to post structured, readable output in a consistent format also required careful prompt engineering in each skill file.
Accomplishments We're Proud Of We are proud that Sprint Guardian actually works. Four agents run in sequence, each building on the previous agent's output, delivering a structured actionable review posted directly on the MR. We are especially proud of the skill file architecture, giving each agent a full identity, purpose statement, guardrails, and self check criteria goes far beyond a basic prompt. Each agent knows what it is, why it exists, and how to verify its own work before posting.
What We Learned The quality of an agent's output is directly proportional to how well it understands its own role — not just its instructions. Agents with identity, purpose, and self check criteria perform dramatically better than agents with only a task description. This reinforced everything we believe about how AI agents should be built.
What's Next for Sprint Guardian Sprint Guardian is the surface layer. Future versions will include:
Adaptive routing based on risk level Team specific rule sets loaded via skills Integration with security scanning tools Compliance reporting layer for enterprise DevSecOps pipelines Sprint Guardian as a standalone product for development teams
Built With
- agent
- anthropic
- architecture
- biz
- claude
- framework
- identity
- javascript
- mke
- multi
- node.js
- orchestration
- skill.md
- stagedintelligence
- tech
