πŸ•

GuardDog

Check if a token is SAFE before trading β€” one call, PASS / WARN / BITE.

The on-chain guard dog for AI agents. Sniffing honeypots, rug pulls & scams before your agent touches them. ι“ΎδΈŠηœ‹ι—¨η‹— 🦴

P95 23ms3 chains: ETHΒ·BSCΒ·SOL4 data sources100 checks/day freeNo signup to try
one call β€” agent-readable answer
$ curl "https://api.guarddog.top/v1/check?chain=bsc&target=0x..." # X-GuardDog-Token: optional during free beta { "verdict": "BLOCK", // BITE! 🦴 do not trade "risk_score": 0.97, "reasons": ["sell_tax 99%", "pool age 2d", "honeypot flagged"], "covered": true, "latency_ms": 23 }

One verdict. Three meanings.

Agents don't read 35-field dashboards. They need an answer they can act on β€” instantly.

βœ“

PASS

Sniffed it. Nothing smells wrong. Safe to proceed.

πŸ‘ƒ

WARN

Sniffed something off. Read reasons, proceed with care (half-size entry).

🦴

BITE! (BLOCK)

Honeypot / rug pattern detected. The dog bites β€” do not trade.

Built agent-first

"The winner in this space will look nothing like a human product."

πŸšͺ Easy to find

MCP servers directory + GitHub + llms.txt + AGENTS.md. Every doorway has the same one-liner and a copy-paste example.

⚑ Zero-friction trial

No signup. 100 free checks/day per IP during beta. Default params are the best params.

πŸ›  Self-explaining errors

Every 4xx carries a hint your agent can execute. Chain aliases normalized (sol→solana) so common mistakes never happen.

πŸ“Š Deterministic answers

verdict + risk_score + reasons array. Same input, same output. Your agent can trust it.

πŸ”¬ Battle-tested

Validated against an ACM dataset of 116,304 labeled Solana rug pulls β€” 100% data coverage, 46.5% flagged suspicious.

πŸ• On duty 24/7

Watching new launches on pump.fun & friends. Daily threat feed on X @OnchainGuardDog.

Quickstart

Two ways in. Both take under a minute.

# 1 β€” REST (curl / any HTTP client)

curl "https://api.guarddog.top/v1/check?chain=eth&target=0x..." \
  -H "X-GuardDog-Token: free-during-beta"

# 2 β€” MCP (Claude / Hermes / any MCP agent)

{ "mcpServers": { "guarddog": { "command": "npx",
  "args": ["-y", "guarddog-mcp"] } } }

# full docs for agents

https://guarddog.top/llms.txt   Β·   https://guarddog.top/AGENTS.md