Skip to content
MythologIQ-Labs-LLCPublic

About

No description, website, or topics provided.

Resources

Code of conduct

Contributing

Security policy

Stars

6 stars

Watchers

0 watching

Forks

Latest commit

 

History

1,994 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

FailSafe

Know what your AI coding agents are doing. Keep development decisions accountable.

Local-first governance, workspace monitoring, and evidence for AI-assisted software development.

VS Code Marketplace Open VSX Latest Release License

Developing FailSafe? Start with the Architecture Index.

Understand and govern AI-assisted development

AI coding agents can write code faster than teams can always review their decisions. FailSafe adds a local, inspectable governance layer to your editor workflow. It helps you see what changed, understand relevant risks, establish intent, and retain evidence for review.

FailSafe is a VS Code extension with a Monitor and a broader Console. It does not promise that every agent action is intercepted, that arbitrary processes are sandboxed, or that VS Code editor saves can be prevented. Controls depend on the action path and integration involved.

A recent call is not end-to-end qualification or proof of customer value. An integration may be configured or successfully contact its service without proving the entire workflow works. Every advertised action needs independent host and provider evidence. FailSafe's internal governance MCP service uses an in-process stdio transport and is not reachable by external MCP clients. A supported agent workflow is not proof that every agent write is intercepted.

What you can do with FailSafe

  • See workspace activity. Review Sentinel findings, change signals, and operational risk in the Monitor and Console.
  • Work from explicit intent. Associate work with an active intent and review policy decisions and higher-risk approvals.
  • Review supported agent workflows. Inspect governed changes and apply them deliberately, rather than assuming every agent write is automatically safe.
  • Trace decisions and evidence. Use checkpoints, audit history, and governance records to understand what happened and what remains unverified.
  • Bring external work into governance. Opt-in Jira and Linear issue previews can become reviewed local intent drafts. Other integrations offer risk ingestion, agent observations, or notifications, depending on their capabilities.

See it in action

FailSafe Console overview

FailSafe Monitor sidebar

For more screens, see the UI and technical reference.

Get started

  1. Install FailSafe from the VS Code Marketplace or Open VSX.
  2. Open a project workspace in VS Code or Cursor.
  3. Open the Console with FailSafe: Open Console (Browser Popout) or Ctrl+Alt+F.
  4. Inspect your workspace with FailSafe: Audit Current File.
  5. Configure governance with FailSafe: Set Up Agent Governance if you want supported agents to receive FailSafe guidance.

The extension's full getting-started and command reference explains optional setup and platform constraints.

Local-first by design

FailSafe's core monitoring and governance run locally. External integrations are opt-in and may send data to their configured providers when you invoke them. Review each integration's privacy and permission contract before enabling it. Local-first is not a claim that every optional integration is offline.

Current release and what's coming

Current Release: v6.0.13 (release notes); the previous release is v6.0.12. See the release history for fixes, security corrections, and packaged platform support.

v6.0.13 highlights: the Integrations Catalog now includes a Statewright AI website handoff. Open the FailSafe Console, go to Integrations → Catalog, find Statewright AI, and select Open Statewright Run History. Statewright handles sign-in and run data in your browser; FailSafe does not embed the history or enforce Statewright workflows. G4 release-ledger witnessing is repository tooling, not an extension feature. Read the branded v6.0.13 release-notes PDF and full changelog.

Newer work on main may not yet be available in the marketplace release. Check GitHub Releases before assuming a repository feature is installed.

For developers and agentic contributors

Start here: Architecture Index. It routes you to canonical design documents, existing decisions, related implementation contracts, and verification. Search that index before adding a new architecture document or changing a design boundary.

Working on FailSafe itself in VS Code? Set the local workspace setting failsafe.architecture.indexPath to ARCHITECTURE_INDEX.md to make Workspace → Architecture read this root index. The extension's general default remains docs/ARCHITECTURE_INDEX.md for other repositories; without the override, this repository's root index may appear as present elsewhere. Local .vscode/settings.json is intentionally ignored, so no contributor setting or installed-user preference is silently changed.

FailSafe is open source under the Apache 2.0 license.

About

No description, website, or topics provided.

Resources

Code of conduct

Contributing

Security policy

Stars

6 stars

Watchers

0 watching

Forks

Releases

Sponsor this project

Packages

Used by

Contributors

Languages