Skip to content
View yashkhou's full-sized avatar

Achievements

Achievement: Pull Sharkx2Achievement: QuickdrawAchievement: YOLO

Achievements

Achievement: Pull Sharkx2Achievement: QuickdrawAchievement: YOLO

Highlights

Block or report yashkhou

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. Markdown is supported. 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
yashkhou/README.md

Yash Khoury: the proof layer for AI agents

Agents say “done”. I build the tools that check.

I'm Yash. I build open-source infrastructure that sits between an AI agent and the real world. It records what the agent did, checks what it claims, and pins down what it's allowed to call. Everything is local-first and model-agnostic, built on boring formats you can inspect.

yashkhou.com · @yashkhou on X

The proof stack

Project What it proves
RunLedger What the agent actually did. Tool calls, decisions and retries are hash-chained, so any edit, deletion or reordering breaks verification.
BrowserProof That the browser really ended up where the agent claims. Explicit assertions, evidence hashes, CI exit codes.
ActionMesh What the agent is allowed to call. One typed contract per capability, served over MCP-style, HTTP and CLI.
Verify That the output is right. Checks DOCX/XLSX/PPTX/PDF files, repos and reversible actions without trusting the generator.
Agent Compat Lab That one repo tells Codex, Claude Code, Gemini CLI and OpenCode the same thing. Outputs SARIF.
Agent Reliability Lab Twelve deterministic tools for hardening agent infrastructure: MCP chaos, contract fuzzing, context firewalls, replay, evals and more. 
npm install github:yashkhou/runledger && npx runledger verify

Also building

  • Glyph: an experimental design language with memory, so agents can reason about structure instead of pixels.
  • Commander Plus: a local-first agent workstation with MCP workspaces, reusable skills, browser control and persistent project context.
  • OpenRetention: self-hosted customer-success software with explainable health scores and revenue-at-risk prioritisation.

Recent upstream work

  • Opened Stellar-agentic #429 to repair repository-specific README links.
  • Reviewed pydantic-ai #8969, a regression fix preventing shared StructuredDict schema metadata from leaking between output types.
  • Added current-main implementation analysis to MCP Python SDK #1933 around stdio ownership and process-stream lifecycle.

If one of these saves you a bad night, a ⭐ helps the next person find it.

Pinned Loading

  1. actionmesh actionmesh Public

    Typed TypeScript action contracts for AI tools, MCP-style calls, HTTP and CLI with shared Zod validation, errors and retry semantics.

    TypeScript

  2. agent-compat-lab agent-compat-lab Public

    Cross-agent compatibility test harness for repository instructions, skills, MCP configuration, and AI coding-agent CI.

    JavaScript

  3. agent-reliability-lab agent-reliability-lab Public

    Twelve deterministic tools for testing and hardening AI agent infrastructure: MCP chaos, contract fuzzing, context firewalls, replay, evals.

    Python

  4. browserproof browserproof Public

    Verify AI browser agents and Playwright workflows with explicit assertions, evidence hashes, HTML reports and CI-ready pass/fail results.

    TypeScript

  5. runledger runledger Public

    Tamper-evident local execution ledger for AI agents: hash-chained tool calls, decisions, retries, results and verifiable HTML reports.

    TypeScript

  6. verify verify Public

    Verification infrastructure for AI work: artifacts, agent-written code, and reversible actions.

    JavaScript