# I built a human approval layer for AI agent payments — looking for MCP testers

**URL:** <https://forum.cursor.com/t/i-built-a-human-approval-layer-for-ai-agent-payments-looking-for-mcp-testers/171326>\
**Category:** Built for Cursor\
**Tags:** auto-run, mcp, security\
**Created:** [September 11, 2026, 8:35am UTC](https://forum.cursor.com/t/i-built-a-human-approval-layer-for-ai-agent-payments-looking-for-mcp-testers/171326 "2026-09-11T08:35:14Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![richard7463](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/richard7463/32/120559_2.png) [@richard7463](https://forum.cursor.com/u/richard7463)\
**Post date:** [September 11, 2026, 8:35am UTC](https://forum.cursor.com/t/i-built-a-human-approval-layer-for-ai-agent-payments-looking-for-mcp-testers/171326/1 "2026-09-11T08:35:14Z")

</div>

I wanted an agent to handle routine actions without giving it unlimited authority.

**AskGrokWallet evaluates each proposed action against a human-written policy:**

**- allow — routine action can continue**

**- ask — pause and wait for a human**

**- deny — do not execute**

Every outcome gets an Ed25519-signed receipt. The receipt records the agent, action, amount, destination, verdict, human

decision and, after settlement, the transaction hash. Receipt events are hash-chained, with the chain head anchored on Ethereum

Sepolia.

The guarded testnet flow now works end to end:

Grok Bot / MCP request → policy → approval inbox → Sepolia vault execution → independently verifiable receipt

Watch the walkthrough and try the no-signup demo:

> **[AskGrokWallet — a wallet with rules for your bots](https://askgrokwallet.io/demo)**
>
> Small things run. Big things ask. Every spend gets a receipt. The approval layer for Grok bots and AI agents.

Verify an anchored example yourself:

> <https://github.com/richard7463/askgrokwallet/blob/main/spec/QUICKSTART.md>

I’m looking for five Grok Bot or Cursor builders to test the MCP workflow. The feedback I need is not “looks cool,” but:

- Which real action would you put behind this gate?

- What context would be missing when you approve it?

- Which part of the receipt would you not trust?

- Where does the MCP workflow become confusing or cumbersome?

Important limitation: this is an unaudited developer preview. The guarded end-to-end settlement has been exercised on Ethereum

Sepolia with mock assets; it has not protected production funds.

---

<div class="post-metadata">

**Author:** ![discourse\_ai\_spam](https://avatars.discourse-cdn.com/v4/letter/d/c68b51/32.png) [@discourse\_ai\_spam](https://forum.cursor.com/u/discourse_ai_spam)\
**Post date:** [September 11, 2026, 8:35am UTC](https://forum.cursor.com/t/i-built-a-human-approval-layer-for-ai-agent-payments-looking-for-mcp-testers/171326/2 "2026-09-11T08:35:18Z")

</div>



---

<div class="post-metadata">

**Author:** ![system](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/system/32/99069_2.png) [@system](https://forum.cursor.com/u/system)\
**Post date:** [September 11, 2026, 8:36am UTC](https://forum.cursor.com/t/i-built-a-human-approval-layer-for-ai-agent-payments-looking-for-mcp-testers/171326/4 "2026-09-11T08:36:55Z")

</div>



---

<div class="post-metadata">

**Author:** ![Joseph\_Zaki](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/joseph_zaki/32/120874_2.png) [@Joseph\_Zaki](https://forum.cursor.com/u/Joseph_Zaki)\
**Post date:** [September 14, 2026, 9:23pm UTC](https://forum.cursor.com/t/i-built-a-human-approval-layer-for-ai-agent-payments-looking-for-mcp-testers/171326/5 "2026-09-14T21:23:42Z")

</div>

Congrats on getting this out Richard!

I tried the demo and ran your receipt verifier. The sample checked out, and changing the signed fields was caught. I found one interesting edge case: a simulated pending anchor response (`blockNumber: null`) still gets marked “fixed onchain” at block 0. Should that stay pending until it lands in a block? Happy to share the small repro.
