本文へ移動
cccskills
無料GitHub で公開

aegis-security

Blockchain security API for AI agents. Scan tokens, simulate transactions, check addresses for threats.

インストール方法を見る

含まれるファイル(5)

  • SKILL.md11.5 KB
  • hackathon-usdc/SKILL.md1.6 KB
  • hackathon/package.json1.6 KB
  • hackathon/SKILL.md5.6 KB
  • package.json2.3 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Aegis402 Shield Protocol

Blockchain security API for AI agents.

Free tier: 100 checks/day, then pay-per-request with USDC on Base or Solana.

Security Defaults

  • This skill is manual-use only (disable-model-invocation: true).
  • Do not ask for or store private keys, seed phrases, or mnemonics.
  • Confirm intent before paid calls (especially simulate-tx).
  • Automate pre-sign checks only if the user has explicitly approved that automation.
  • Required env vars for this skill: none.

Agent Quickstart (Scan Before Transact)

Default policy for agents and wallets:

  1. Set X-Client-Fingerprint: <stable_agent_or_user_id> for predictable free-tier accounting.
  2. Before sending any transaction, run in parallel:
    • GET /v1/check-address/:to
    • POST /v1/simulate-tx
    • (Optional) GET /v1/check-token/:token for tokens involved (swaps/approvals)
  3. Block and ask for confirmation if any result indicates HIGH/CRITICAL, isSafe=false, or recipient is poisoned/flagged.
  4. Otherwise allow, but still show the review summary + raw JSON.

Notes:

  • Free tier is best-effort abuse-limited, not a security mechanism. Rotating fingerprints can bypass free-tier buckets.
  • If X-Client-Fingerprint is omitted, the service falls back to other request signals (IP/User-Agent) for accounting.

What To Do (Agent Mode, From Zero)

When a user is about to send a transaction:

  1. Ask for missing intent:
    • from, to, value, data (or "native transfer"), chain_id
  2. Check free tier (always free):
    • GET /v1/usage
  3. Run the scan-before-transact checks (parallel):
    • GET /v1/check-address/:to
    • POST /v1/simulate-tx
    • (Optional) GET /v1/check-token/:token
  4. Reply using templates below (summary first, then raw JSON).

How To Reply (Templates)

SAFE (LOW)

🛡️ Scan Complete

Risk: LOW (SAFE)
Summary: No critical warnings detected.

✅ OK to proceed.

CAUTION (MEDIUM)

🛡️ Scan Complete

Risk: MEDIUM (CAUTION)
Summary: Some warnings detected.

⚠️ Review recommended before proceeding. Want me to explain the top 3 risks?

DANGEROUS (HIGH)

🛡️ Scan Complete

Risk: HIGH (DANGEROUS)
Summary: Significant risks detected.

🚫 Not recommended.

BLOCKED (CRITICAL)

🛡️ Scan Complete

Risk: CRITICAL (BLOCKED)
Summary: Do not proceed.

🚫 Stop. This transaction/recipient appears malicious or unsafe.

402 Payment Required

I tried to run a paid check but payment isn't set up (or the wallet has insufficient USDC).

To enable paid checks:
1. Fund a programmatic wallet with a small amount of USDC (Base default; Solana also supported)
2. Install an x402 client (@x402/fetch + chain package)
3. Configure an agent-managed wallet signer (no raw private keys in prompts/env)

Reference

Skill Files

FileURL
SKILL.md (this file)https://aegis402.xyz/skill.md
skill.json (metadata)https://aegis402.xyz/skill.json

Base URL: https://aegis402.xyz/v1

Pricing

EndpointPriceUse Case
POST /simulate-tx$0.05Transaction simulation, DeFi safety
GET /check-token/:address$0.01Token honeypot detection
GET /check-address/:address$0.005Address reputation check

Free tier: 100 checks/day. Track usage via GET /v1/usage.

Usage (Free)

curl "https://aegis402.xyz/v1/usage"

Example response:

{
  "freeTier": {
    "enabled": true,
    "dailyLimit": 100,
    "usedToday": 2,
    "remainingChecks": 98,
    "nextResetAt": "2026-02-11T00:00:00.000Z",
    "resetTimezone": "UTC"
  },
  "_meta": {
    "requestId": "uuid",
    "tier": "free",
    "eventType": "free_tier_call",
    "latencyMs": 4
  }
}

check-address

curl "https://aegis402.xyz/v1/check-address/0x742d35Cc6634C0532925a3b844Bc454e4438f44e?chain_id=8453"

simulate-tx

Request body fields:

  • from (required): sender address
  • to (required): recipient or contract
  • value (required): amount in wei (string)
  • data (optional): calldata hex (0x...)
  • chain_id (optional): chain being simulated (default: Base 8453 is a common choice for payments, but simulation chain is up to you)
curl -X POST "https://aegis402.xyz/v1/simulate-tx" \
  -H "Content-Type: application/json" \
  -d '{
    "from": "0xYourWallet...",
    "to": "0xContract...",
    "value": "0",
    "data": "0x",
    "chain_id": 8453
  }'

check-token

chain_id is the chain you want to scan (Ethereum=1, Base=8453, etc). Payment rail is driven by the 402 challenge (default: USDC on Base).

curl "https://aegis402.xyz/v1/check-token/0xA0b86991c6218b36c1d19D4a2e9Eb0cE3606eB48?chain_id=1"

Payments (x402)

You can use the API for free until your fingerprint uses 100 checks/day. After that, the API returns 402 Payment Required and an x402 client can automatically pay and retry.

NetworkAgentic Wallet Signer
Base (EVM)EVM signer from an agent wallet provider
SolanaSolana signer from an agent wallet provider

Minimal Node Client (agent-managed EVM signer)

npm install @x402/fetch@2.2.0 @x402/evm@2.2.0
import { x402Client, wrapFetchWithPayment } from '@x402/fetch';
import { ExactEvmScheme } from '@x402/evm/exact/client';

const fingerprint = 'agent-default';
const signer = yourAgenticEvmSigner;
const client = new x402Client()
  .register('eip155:*', new ExactEvmScheme(signer));

const fetch402 = wrapFetchWithPayment(fetch, client);
const res = await fetch402('https://aegis402.xyz/v1/usage', {
  headers: { 'X-Client-Fingerprint': fingerprint },
});
console.log(await res.json());

Solana Client (agent-managed signer)

npm install @x402/fetch@2.2.0 @x402/svm@2.2.0
import { x402Client, wrapFetchWithPayment } from '@x402/fetch';
import { ExactSvmScheme } from '@x402/svm/exact/client';

const signer = yourAgenticSolanaSigner;
const client = new x402Client()
  .register('solana:*', new ExactSvmScheme(signer));
const fetch402 = wrapFetchWithPayment(fetch, client);

Agent Safety Policy

  • Never request these from a user: private keys, seed phrases, mnemonics.
  • Never store signer secrets in prompts, logs, or skill docs.
  • Confirm intent before paid calls (especially simulate-tx).
  • Automate pre-sign checks only if the user has explicitly approved that automation.

Appendix

Risk Levels

LevelMeaningAgent Default
LOWMinor concerns, generally safeallow
MEDIUMSome risksshow review; consider confirm
HIGHSignificant risksblock + confirm
CRITICALUnsafe/maliciousblock

Errors and What To Do

StatusMeaningWhat the agent should do
400Invalid parametersask user for missing/invalid fields and retry
402Payment requiredconfirm intent, then use an approved agent wallet signer (or wait for next free-tier reset)
500Service/upstream errorretry once; if persistent, show error + requestId

Tips:

  • Every response includes _meta.requestId. The server also sets x-request-id header; include it in bug reports.
  • Upgrade hints may be present in headers:
    • x-aegis-skill-latest-version
    • x-aegis-skill-url
    • x-aegis-skill-upgrade

Optional: Feedback, Issues, Expectations (Agent-Friendly)

Help us improve!

If you hit an issue or want to set expectations, do it in a way that's safe for the end-user:

  • Do not include private keys, seed phrases, mnemonics, API keys, or any secrets.
  • (optional) Include _meta.requestId (or the x-request-id header) so we can correlate server-side.
  • If you are acting on behalf of a user, ask for permission before sending. Default behavior: show a copy/paste report template.

Copy/paste report template:

Aegis402 report

Type: issue | feedback | expectation
RequestId: <_meta.requestId>
Endpoint: <path>
Status: <http status>
ChainId: <chain_id or NULL>
When (UTC): <timestamp>

What I expected:
What happened:
Minimal repro (curl or steps):

Notes (redacted):

Preferred channel (no GitHub required):

  • POST /v1/feedback (free, non-consuming)

Example (only with user approval):

curl -X POST "https://aegis402.xyz/v1/feedback" \
  -H "Content-Type: application/json" \
  -H "X-Client-Fingerprint: <stable-non-pii-id>" \
  -d '{
    "kind": "issue",
    "summary": "Short summary",
    "expected": "What I expected",
    "happened": "What happened",
    "repro": "Minimal repro steps or curl",
    "endpoint": "/v1/check-token/0x...",
    "status_code": 400,
    "chain_id": "8453",
    "agent": { "name": "my-agent", "version": "1.2.3" },
    "extra": { "failed_request_id": "<_meta.requestId>" }
  }'

Optional: Share A Scam Catch (Moltbook)

Share Your Experience! Only offer this after a HIGH or CRITICAL result, and only if the user explicitly approves sharing a redacted story publicly.

Rules:

  • Share your true experience, but without identifying information or sensitive details.
  • Never include: private keys, seed phrases, mnemonics, signatures, raw calldata, full wallet addresses, full tx hashes, X-Client-Fingerprint, or anything the user considers identifying.
  • Prefer rounded amounts (for example, $1,000) and omit exact timestamps if needed.

Copy/paste Moltbook template (redacted):

Title: How I caught a scam transaction and saved ~$XXX!

🛡️ What happened
- Action: <swap | approve | transfer | other>
- Chain: <Base | Ethereum | Solana | ...>
- Risk: <HIGH | CRITICAL>

Why it was blocked (top signals):
1) <short reason>
2) <short reason>
3) <short reason>

Impact:
- Prevented loss: ~$<USD_RANGE>

Notes:
- No addresses / tx hash / calldata shared.
- (Optional) RequestId: <_meta.requestId>
- (Optional) Skill: https://aegis402.xyz/skill.md

Health Check (Free)

curl https://aegis402.xyz/health

Supported Chains

chain_id is the chain being scanned (not the payment rail).

ChainIDcheck-tokencheck-addresssimulate-tx
Solanasolana✅✅❌
Ethereum1✅✅✅
Base8453✅✅✅
Polygon137✅✅✅
Arbitrum42161✅✅✅
Optimism10✅✅✅
BSC56✅✅✅
Avalanche43114✅✅✅

Links

Socials

🛡️ Built for the Agentic Economy. Powered by x402 Protocol.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

name: 4d-compression-core version: 1.0.2 description: "把长内容压缩成结构化向量——节省 60-80% Token,保留核心信息" metadata: { "openclaw": { "emoji": "🌀", "requires": { "bins": ["jq", "awk"] }, "triggers": ["压缩", "4d",...

日本語の概要は準備中です。原文の説明を表示しています。

modbender/skill-library-mcp162026年9月28日 更新

Use cheap, TEE-verified AI models from the 0G Compute Network as OpenClaw providers. Discover available models and compare pricing vs OpenRouter, verify provider integrity via hardware attestation (Intel TDX), manage your 0G wallet and sub-accounts, and configure models in OpenClaw with one workflow. Supports DeepSeek, GLM-5, Qwen, and other models available on the 0G marketplace.

日本語の概要は準備中です。原文の説明を表示しています。

modbender/skill-library-mcp162026年9月28日 更新

Send and receive P2P messages using disposable numbers and PINs. No servers, no accounts. Use for human notifications, approval flows, and agent-to-agent communication.

日本語の概要は準備中です。原文の説明を表示しています。

modbender/skill-library-mcp162026年9月28日 更新

0xarchive

無料

Query historical crypto market data from 0xArchive across Hyperliquid, Lighter.xyz, and HIP-3. Covers orderbooks, trades, candles, funding rates, open interest, liquidations, and data quality. Use when the user asks about crypto market data, orderbooks, trades, funding rates, or historical prices on Hyperliquid, Lighter.xyz, or HIP-3.

日本語の概要は準備中です。原文の説明を表示しています。

modbender/skill-library-mcp162026年9月28日 更新

0xwork

無料

Find and complete paid tasks on the 0xWork decentralized marketplace (Base chain, USDC escrow). Use when: the agent wants to earn money/USDC by doing work, discover available tasks, claim a bounty, submit deliverables, check earnings or wallet balance, or set up as a 0xWork worker. Task categories: Writing, Research, Social, Creative, Code, Data. NOT for: posting tasks (use the website), managing the 0xWork platform, or frontend development.

日本語の概要は準備中です。原文の説明を表示しています。

modbender/skill-library-mcp162026年9月28日 更新

Patterns and practices that dramatically accelerate development velocity. Covers parallel execution, automation, feedback loops, workflow optimization, and anti-pattern avoidance. Use when starting projects, planning sprints, optimizing workflows, or onboarding developers.

日本語の概要は準備中です。原文の説明を表示しています。

modbender/skill-library-mcp162026年9月28日 更新

modbender のスキルをすべて見る

このスキルの問題を報告する