Open, human-governed assurance for AI-assisted engineering.
Agile V helps teams turn AI-assisted work into a reviewable chain from approved requirements to implementation, tests, independent verification, and human release decisions. It is a tool-agnostic operating and evidence layer for coding agents, engineering teams, and controlled delivery workflows.
Intent -> reviewed requirement baseline -> build + independent test design
-> independent verification -> human release decision
| Control | Practical effect |
|---|---|
| Baselined requirements | Drafts and chat context do not silently become build input |
| Typed traceability | Requirements, artifacts, tests, risks, approvals, and evidence retain explicit lineage |
| Independent challenge | The builder does not verify its own work |
| Human Gates | Accountable people approve requirements, exceptions, residual risk, and release |
| Risk-scaled rigor | L0-L4 selects proportionate controls and evidence |
| AI provenance | Material model, runtime, tool, context, and artifact influence can be recorded without hidden reasoning or secrets |
Agile V does not establish certification, regulatory approval, legal compliance, product safety, or organizational conformity. It supports engineering and governance processes; qualified people remain responsible for applicability, operation, evidence review, and decisions.
The AgentSkills.io-compatible contract library for Claude Code, Cursor, VS Code, GitHub Copilot, and other compatible agents.
- 45 cataloged skills across requirements, implementation, verification, validation, safety, compliance, release, observability, and existing-repository analysis
- Five implementation domains: Python, JavaScript/TypeScript, NestJS, Dart/Flutter, and embedded C/C++
- 18 machine-readable evidence schemas and 348 deterministic repository tests
- Current release: v3.9.0
Start in 60 seconds · Choose a profile · Browse tutorials
A Python scaffold for applying Agile V controls in a repository with structured briefs, evidence bundles, traceability, and CI gates.
Explore the scaffold · Read its documentation
| Goal | Resource |
|---|---|
| Understand the lifecycle | Golden Journey |
| Select skills | Installation Profiles and Routing Guide |
| Try a concrete workflow | Practical Tutorials |
| Review assurance boundaries | Compliance Posture |
| Ask a question | GitHub Discussions |
| Report a defect or propose a skill | Issues |
Contributions, reproducible issue reports, bounded case studies, and critical feedback are welcome. Start with the contribution guide, showcase guide, and security policy.
agile-v.org · info@agile-v.org · GitHub Discussions
The public Agent Skills library is licensed under CC BY-SA 4.0. Repository-specific licenses apply to other projects.