Skip to content
View rsh1k's full-sized avatar

Block or report rsh1k

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rsh1k/README.md

I build governance and security tooling for AI agents.

A complete product, self-hosted Go binary with no dependencies, under Apache-2.0:

  • Quilzo — the control plane for AI agents and the content they produce: every agent gets an identity, a policy, a budget and a signed record, and a person approves the actions that matter · manual · quilzo.io
  • aertun — the wind tunnel for AI agents: scripted malicious agent sessions replayed through your own telemetry, scored honestly, the scorecard signed and verifiable by anyone · manual

And six Apache-2.0 packages that work alone and compose into one control plane:

  • revoco — undo for AI-agent actions: the rollback is planned before the action runs, proven by drills, and a compromised grant's whole blast radius rolls back in one call
  • retie — a terminal coding agent that classifies every action by how hard it is to undo
  • mcp-gate — a policy firewall for MCP tool calls, with redaction on the wire
  • veritrail — tamper-evident provenance: prove who authorized any agent action
  • mnemosyne — a memory-integrity firewall for agents that remember
  • irredux — route AI agent actions by how hard they are to undo, not by how hard they look.

Start at agent-governance — the 60-second tour with measured numbers.

@rsh1k's activity is private