control-plane

control-plane is an agent for Claude Code from 0xSoftBoi/suwappubot. It costs 0 tokens per session (7,658 once invoked), scanned C, original, Apache-2.0.

A policy checkpoint for an agent that can move funds, requiring approved rules to be evaluated before a transaction is built or sent.

In plain words
What is it for?
Use it to control swap requests made through API endpoints or the execute-swap MCP tool, with policies scoped to an organization or agent.
Why use it?
It prevents fund-moving requests from reaching transaction execution when they do not satisfy the configured policies, while leaving read-only requests ungated.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md).

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/0xsoftboi/suwappubot/control-plane
Clone the repo
git clone --depth 1 https://github.com/0xSoftBoi/suwappubot

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for control-plane

README.md
[![agentmods](https://agentmods.dev/badge/agents/0xsoftboi/suwappubot/control-plane.svg)](https://agentmods.dev/agents/0xsoftboi/suwappubot/control-plane)
Your own site
<a href="https://agentmods.dev/agents/0xsoftboi/suwappubot/control-plane"><img src="https://agentmods.dev/badge/agents/0xsoftboi/suwappubot/control-plane.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 7,658 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00000 $0.07658
Opus 5 $0.00000 $0.03829
Sonnet 5 $0.00000 $0.01532
Haiku 4.5 $0.00000 $0.00766

Measured 6d ago against content hash eed7c19e3fcc, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade C, and why

control-plane scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Cloud metadata endpointhighServer-side request forgery

One request to 169.254.169.254 can return temporary IAM credentials.

`169.254.169.254` — defends against DNS rebinding, not just a static
docs/agents/control-plane.md · 585 lines

How it starts

The opening of the file, as written. The whole thing — 585 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Suwappu Agent Control Plane — Integrator Guide

This document describes the control plane as actually shipped on this branch (feat/agent-control-plane-v2). An earlier version of this file described a superseded design (agent_approvals table, intent-hash + 5% value band, /webapp/approvals endpoints) — that design was replaced by the approval_requests payload-hash flow documented below. Do not conflate the two; the old table/fields no longer exist in this branch's schema.

1. Overview

The control plane gates fund-moving agent calls (POST /v1/agent/swap, POST /v1/agent/swap/execute, and the MCP execute_swap tool) through a shared evaluator, PolicyService.evaluate(), before any transaction is built or broadcast. Read-only MCP tools (quotes, portfolio, prices, etc.) are not policy-gated — only execute_swap.

  • Policies (api-ts/src/db/schema/policies.ts, policies table, evaluated by api-ts/src/services/PolicyService.ts) are rows scoped to an org (organizationId), optionally further narrowed to one agent (agentId set), or scoped to a bare agent with no org at all (organizationId null, agentId set — a plain agent-token/MCP-auth request with no org API key). Fields: maxTxUsd, maxSlippageBps, maxGasUsd, dailyCapUsd, sessionCapUsd, maxTxPerHour, allowedChains/blockedChains, allowedTokens/blockedTokens, destinationAllowlist, allowedContracts (router/contract allowlist — without a matched contractAddress on the intent this gate enforces nothing), requireApprovalAboveUsd, approvalMode ('above_limit' default | 'always_ask' | 'autonomous'), expiresAt (time-boxed grant — an expired policy is skipped entirely), priority (lower wins first), enabled.
  • Kill switchespolicy_kill_switches, three scopes: global, org, agent. Any active match short-circuits to block, checked before ordinary policy rows. Kill-switch reads fail CLOSED (DB unreadable → block); ordinary policy reads fail OPEN (DB unreadable → allow, logged) — see §8 for why this matters for migration ordering.
  • Approval loop — a require_approval verdict inserts a pending approval_requests row. The resolved human owner (see §3's owner resolution) approves/denies it; the agent then re-submits the original request with approval_id set.
  • Audit chainaudit_logs, hash-chained (entryHash/prevHash), append-only. Every non-allow policy decision, approval create/approve/deny, link/unlink, and kill-switch change writes a row. Verifiable via GET /v1/agent/audit/verify.

Read the full file on GitHub · 585 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 6d ago First seen · 585 lines · 0 tokens per session scan C eed7c19e3fcc

Subscribe to this mod's changes

control-plane is an agent published in the GitHub repository 0xSoftBoi/suwappubot (3 stars, last pushed today), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 7,658 tokens. A static security scan graded it C with 1 finding (cloud metadata endpoint). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories

defi-trader

Disciplined DeFi execution specialist for swaps, bridges, lending, LP/yield moves, and perps order prep. Use when the user wants to actually DO something on-chain. Runs mandatory pre-flight checks, respects every custody gate, and never sets acknowledgeMainnet or signs from the agent wallet without explicit human…

ChainGPT-org/chaingpt-claude-skill · 74 tokens

yield-aggregator-specialist

Yield aggregator and ERC-4626 specialist. Yearn V3, Beefy, Sommelier, MetaMorpho, custom vaults with strategies. Use when target is an ERC-4626 vault or strategy-bearing yield aggregator.

omermaksutii/RugProof · 54 tokens

staking-specialist

Staking-protocol specialist. Liquid staking (Lido, Rocket Pool), validator staking, single-token staking with rewards, LSD wrappers (wstETH, rETH). Use when target is a staking contract.

omermaksutii/RugProof · 46 tokens

chainaware-portfolio-risk-advisor

Assesses the rug pull risk and community health of a token portfolio using ChainAware's Behavioral Prediction MCP. Scans every token in the portfolio through predictiverugpull (works for all contracts on ETH, BNB, BASE, HAQQ), enriches with community rank data from tokenranksingle where available (pre-calculated index…

ChainAware/behavioral-prediction-mcp · 290 tokens

contract-auditor

Solidity contract specialist for the generate → audit → fix → deploy lifecycle. Use when writing, reviewing, auditing, or deploying smart contracts. Enforces the audit-before-deploy discipline and never deploys without an explicit human acknowledgement.

ChainGPT-org/chaingpt-claude-skill · 51 tokens

chainaware-wallet-auditor

Specialized Web3 intelligence analyst powered by ChainAware's Behavioral Prediction MCP. Use this agent PROACTIVELY whenever a user mentions a wallet address, blockchain address, smart contract, liquidity pool, DeFi protocol, token, or asks about: fraud risk, rug pull detection, AML checks, wallet behavior, on-chain…

ChainAware/behavioral-prediction-mcp · 139 tokens