# Design GenLayer Supervisor Reviews

Delegated agents need review rules before they can safely keep authority. This skill helps define GenLayer evidence, rubrics, decisions, and revocation handoffs.

## Install

```bash
npx skillstore add internet-court/genlayer-intelligent-contracts
```

## Metadata

- Status: approved
- Slug: internet-court-genlayer-intelligent-contracts
- Skillstore revision: r1
- Version status: missing
- Tree hash: fd0fc1a0017b3af573397d0f4e413f175ad1444b11d82a02190c445f30bb4fc1
- Author: internet-court
- GitHub username: internet-court
- License: MIT
- Repository: https://github.com/internet-court/internet-court-skill/tree/main/integrations/genlayer-intelligent-contracts
- Ref: 3f6e026a3363e0954ede7bef0cfe88d4475de137
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, filesystem
- Quality score: 74
- Public page: https://skillstore.pages.dev/skills/internet-court-genlayer-intelligent-contracts
- Manifest: https://skillstore.pages.dev/api/skills/internet-court-genlayer-intelligent-contracts/manifest

## Capabilities

- Defines review inputs for mandates, action receipts, outputs, costs, and prior decisions.
- Drafts rubrics for relevance, scope compliance, safety, budget discipline, and evidence quality.
- Specifies decision outputs for continue, warn, constrain, revoke, and escalate outcomes.
- Models wallet-spend snapshots for delegated on-chain agent reviews.
- Outlines ERC-7710 revocation handoffs when a connector skill defines the implementation path.

## Use Cases

- Review On-Chain Agents: Define evidence and scoring for agents that spend funds, deploy contracts, or call external protocols.
- Specify Revocation Workflows: Map GenLayer decisions to constrain or revoke outcomes before implementation starts.
- Prepare Supervisor Requirements: Create a structured brief for teams building GenLayer Intelligent Contract supervisors.

## Prompt Templates

### Define a Review Flow

```
Design a GenLayer supervisor review for an agent with this mandate: [mandate]. Include evidence, rubric, decisions, and failure paths.
```

### Build a Rubric

```
Create a rubric for an on-chain agent that can [actions]. Cover safety, scope, cost, evidence, and result quality.
```

### Map ERC-7710 Effects

```
Given this delegation policy and revocation controller plan, specify how constrain and revoke decisions should reach the target chain.
```

### Audit a Supervisor Design

```
Review this GenLayer supervisor design for missing evidence, weak revocation assumptions, bad relayer behavior, and appeal gaps: [design].
```

## Limitations

- Does not write, lint, test, deploy, or interact with GenLayer contracts.
- Does not provide a complete bridge, relayer, or EVM controller implementation.
- Requires external evidence sources for receipts, artifacts, costs, and transaction state.
- Cannot guarantee trustless revocation without a verified cross-chain effect path.

## Best Practices

- Separate judgment logic from the bridge or controller that applies authority changes.
- Require receipts, artifact hashes, costs, and rationales for every reviewed action.
- Test continue, warn, constrain, revoke, missing evidence, and bad relayer paths.

## Anti Patterns

- Do not claim revocation is trustless without a defined verification path.
- Do not treat qualitative GenLayer decisions as automatic EVM state changes.
- Do not omit appeal and emergency revoke paths for long-lived authority.

## Security Audit

- Audited at: 2026-07-21T03:16:23.422\+00:00
- Summary: All 33 static detections are false positives caused by Markdown code fences, TypeScript template-literal types, documentation URLs, and a fixed relative skill reference. The reviewed files contain no executable commands, network operations, prompt-injection language, or evidence of data exfiltration.

## Stats

- Views: 0
- Downloads: 5
- Favorites: 1
- Popularity score: 0