Bringing Chip Design Tools into Modern GitHub‑Based Hardware Development

Achim Nohl, Buvanesh Balasubramanian, Daniel Castelló, Varun Shah

May 27, 2026 / 5 min read

Subscribe to Our Blog
Thanks for subscribing to the blog! You’ll receive your welcome email shortly.

Introduction

Hardware development is increasingly following the path software followed over the last decade. Design teams are adopting Git-based collaboration, CI/CD practices, and GitHub pull requests not just for software, but for chip RTL, verification, and integration flows. At the same time, teams are beginning to introduce agentic workflows – AI and rule-based agents that analyze code, reason for design intent, and take action to accelerate development.

But one gap remains persistent – EDA tools are not built for CI/CD or agent-driven execution. They typically live outside the GitHub workflow, require manual invocation, and have limited programmatic interfaces. As a result, teams experiment with custom connectors, APIs, or MCP servers to allow CI pipelines or agents to trigger these tools.

With Synopsys Cloud's support for GitHub Apps, that gap disappears. "GitHub Apps on Synopsys Cloud let our engineers and agents run Synopsys EDA flows directly from GitHub," said Benjamin Holtzman, Staff Design Enablement Engineer at Ahead Computing. "This brings software style, agile, CI/CD development to hardware design, with EDA flows triggered from GitHub pull requests – no need to switch to a traditional design environment."

Hardware Is Embracing Software Development Practices

Modern hardware teams now work much like software teams:

  • RTL and verification code lives in GitHub repositories
  • Pull requests are the primary review and integration mechanism
  • CI pipelines enforce quality and consistency
  • Automation runs early and often, not just at the end

This shift delivers clear benefits – faster iteration, better collaboration, and earlier defect detection – but only if EDA tools participate directly in these workflows.

Traditional approaches force designers to leave GitHub to:

  • Log into separate systems
  • Run batch jobs
  • Wait for reports
  • Manually share results back into reviews

That breaks the CI/CD loop. Collaboration happens across disconnected systems. Designers run tools in LSF, architects review reports in email threads, CAD managers track license usage in spreadsheets, and Git commits exist separately from verification results.

GitHub Apps: EDA as a Native CI/CD Capability

Synopsys Cloud GitHub Apps bring enterprise EDA workloads directly into GitHub events such as:

  • Pull request creation or updates
  • Code pushes
  • Issue comments or labels
  • Slash command invocations

From the user or pipeline perspective, the experience is simple:

  • Code changes occur in GitHub
  • GitHub Apps get triggered (event-sensitive or command-driven)
  • GitHub Apps run EDA tools automatically
  • Results appear directly in the pull request as checks, comments, or artifacts
Make code edits, run EDA flows and analyze results all within GitHub

Make code edits, run EDA flows and analyze results all within GitHub

No VPNs. No remote desktops. No manual handoffs. EDA becomes CI-native, not an external dependency. This takes collaboration to a new level where every interaction – code changes, agent decisions, EDA tool invocations and results, human review – is captured in a single pull request thread.

The Synopsys Linter GitHub App triggered using a slash command

The Synopsys Linter GitHub App triggered using a slash command

Details of automated fixes made by the Synopsys Linter GitHub App

Details of automated fixes made by the Synopsys Linter GitHub App

The Plumbing That Makes It Work

Enabling GitHub, CI, and agent-driven EDA is not just about exposing tools. It requires robust infrastructure to reliably translate GitHub events into secure, correctly scoped EDA execution.

Synopsys Cloud provides this plumbing out of the box. When a GitHub App is triggered from a pull request (whether by a commit, a comment, or an agent-generated action), Synopsys Cloud handles the entire lifecycle:

  • Securely receives and validates the GitHub event
  • Resolves the context of the request
  • Routes the workload to the correct execution environment
  • Executes the EDA workload as an ephemeral job
  • Returns structured results back to GitHub

All of this happens automatically without users having to build orchestration layers, manage routing logic, or maintain environment-specific connectors.

 

Security-First Architecture with Zero-Trust Options

Semiconductor companies operate under strict IP protection and compliance requirements. GitHub Apps on Synopsys Cloud are designed with security as a first principle.

The platform supports three security models:

  • Common Apps (Shared Keys): Enables zero-setup, rapid deployment, and full managed support, governed by contractual controls.
  • Customer Apps (Synopsys-Managed Keys): Provides strong IP isolation without adding operational overhead.
  • Customer Apps (Customer-Managed Keys): Enabling maximum IP protection and compliance readiness (e.g., ITAR, FedRAMP) for zero-trust or regulated environments

The Rise of Agentic Hardware Development

As teams mature with CI/CD, many are now exploring agentic development – AI-driven or rule-based agents that:

  • Analyze code and design changes
  • Decide which checks or analyses are required
  • Trigger tools
  • Interpret results
  • Propose fixes or follow-up actions

In many architectures, this requires explicit tool connectors, often via MCP servers, custom service layers, and long-running integration gateways. GitHub Apps offer a far simpler and more natural solution even for regulated, on-prem, or zero-trust environments.

 

GitHub as the Agent-to-EDA Connector

Most agentic systems already integrate deeply with GitHub. They can:

  • Read pull requests and diffs
  • Post comments or labels
  • Open or update pull requests
  • Monitor checks and statuses

By exposing EDA tools as GitHub Apps, Synopsys makes EDA directly accessible to agentic flows through GitHub itself. For an agent, invoking an EDA tool becomes as simple as creating a GitHub event.

 

Example Agentic Flow

A typical workflow might look like this:

  1. An agent analyzes a pull request in GitHub
  2. It determines lint, CDC, or formal checks are required
  3. It triggers a Synopsys GitHub App (automatically or via a generated comment)
  4. Synopsys Cloud executes the EDA workload in the correct environment
  5. Results return to GitHub as structured outputs the agent can reason over

From the agent's perspective, EDA is just another GitHub-addressable action.

EDA at the Speed of GitHub

Hardware development is evolving toward software-style workflows, continuous verification, and increasingly autonomous tooling.

GitHub Apps on Synopsys Cloud were designed for teams that:

  • Run hardware development with CI/CD discipline using GitHub
  • Want EDA feedback early and continuously
  • Are experimenting with agentic workflows
  • Need strong security, isolation, and compliance

By using GitHub as the integration surface, Synopsys eliminates the need for bespoke EDA connectors and enables agent-ready EDA workflows with minimal friction, immediately consumable by both humans and agents.

Continue Reading

ASK SYNOPSYS
BETA
Ask Synopsys BETA This experience is in beta mode. Please double check responses for accuracy.

End Chat

Closing this window clears your chat history and ends your session. Are you sure you want to end this chat?