Neurobro Meme Hunting

Neurobro Meme Hunting

NeuroAPI's /agent/ask agent can scan trending DEX activity, surface newly launched tokens, and run on-chain security audits. This skill turns that into a disciplined workflow for finding memecoin candidates and cutting the obvious scams before they reach the user.

Risk reality - read this first

Memecoins are the highest-risk corner of crypto. Most go to zero, and a large share are outright scams - honeypots, rug pulls, and tax traps. This skill helps discover and screen candidates; it does not make any token safe, and nothing it produces is financial advice. Always:

• Treat every result as research, not a recommendation.
• Run the audit step - never surface a discovery without vetting it.
• Tell the user, plainly, that they can lose 100% of what they put in.

When to use this skill

• Finding new or trending memecoins on a chain (Solana, Base, Ethereum, BSC…)
• Screening a specific token before the user buys
• Checking whether a token is a honeypot, a rug risk, or holder-concentrated

Do not use this skill for trade entry/exit planning (use neurobro-trading-setup) or broad market research (use neurobro-market-research).

Calling NeuroAPI

Every request goes to one endpoint:

POST https://api.neurobro.ai/api/v1/agent/ask

Authenticate with the X-API-Key header, reading the key from the NEUROAPI_API_KEY environment variable - never hardcode it. The body takes a prompt and a mode (fast | smart | max, default smart).

curl -s https://api.neurobro.ai/api/v1/agent/ask \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $NEUROAPI_API_KEY" \
  -d '{"prompt": "", "mode": "smart"}' | jq -r .answer

For the full request/response schema, status codes, and rate limits, see the official docs: .

The hunt: discover, filter, audit, rank

Four steps. The audit step is mandatory - discovery without vetting is how users get rugged.

1. Discover

Ask the agent for trending or newly launched tokens. Always name the chain and a time window - memecoin activity is chain-specific and fast-moving.

List the top trending liquidity pools on Solana from the last 24 hours.
For each token give: name, ticker, pool age, liquidity (USD), 24h volume,
24h price change, and transaction count.

For fresh launches specifically, ask for recently added tokens rather than trending ones - the two surface different candidates.

2. Filter

Narrow the list to tokens with real, durable activity. Push the filter into the prompt so the agent does the work. Sensible starting thresholds:

| Signal | Baseline filter | Why | | --- | --- | --- | | Liquidity | $50-100k or more | Thin liquidity is trivial to rug and impossible to exit. | | 24h volume | > liquidity | Real trading, not a static pool. | | Pool age | Match the strategy | . Check for: honeypot / sell restrictions, contract ownership and mint authority, buy and sell tax, liquidity lock or burn status, and top-10 holder concentration. End with a clear risk verdict: avoid, high-risk, or relatively clean.

### 4. Rank and decide

Drop anything that fails the audit (see Red flags below). For survivors,
present the user a short ranked list with: ticker, chain, liquidity, age, the
audit verdict, and the single biggest remaining risk. Never present a winner -
present candidates with their risks attached.

## Recipes

### Screen one token end to end (Python)

```python
import os, httpx

def ask(prompt: str, mode: str = "smart") -> str:
    res = httpx.post(
        "https://api.neurobro.ai/api/v1/agent/ask",
        headers={"X-API-Key": os.environ["NEUROAPI_API_KEY"]},
        json={"prompt": prompt, "mode": mode},
        timeout=60,
    )
    res.raise_for_status()
    return res.json()["answer"]

audit = ask(
    "Audit the token at base:0x1234...abcd - honeypot, ownership, mint "
    "authority, buy/sell tax, LP lock, and top-10 holder concentration. "
    "Give a clear risk verdict."
)
print(audit)

Quick trending scan (curl)

curl -s https://api.neurobro.ai/api/v1/agent/ask \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $NEUROAPI_API_KEY" \
  -d '{"prompt": "Top 10 trending memecoins on Base in the last 12 hours with liquidity over $100k - name, ticker, liquidity, 24h volume, age.", "mode": "fast"}' | jq -r .answer

Red flags - reject on sight

If the audit surfaces any of these, the token fails. Tell the user why:

• Honeypot - buys succeed but sells are blocked or restricted.
• Mint authority active - the deployer can print unlimited new supply.
• Ownership not renounced - the contract can still be changed post-launch.
• High or mutable tax - buy/sell tax above ~10%, or tax the owner can raise.
• Liquidity not locked or burned - the pool can be pulled at any moment.
• Holder concentration - top 10 wallets hold a large share (~30%+) of

supply; a few sells crater the price.

• Liquidity tiny vs. market cap - the "value" is unbacked and illiquid.

Green flags - necessary, never sufficient

Renounced ownership, revoked mint authority, locked/burned liquidity, low and fixed tax, and dispersed holders are the baseline for "not an obvious scam." They do not make a token a good buy - memecoins with clean contracts still routinely go to zero on momentum alone.

Choosing a mode

• fast - quick trending scans where you only need the candidate list.
• smart (default) - the audit and ranking steps; use it for any vetting,

where careful reasoning over the risk signals matters.

• max - deep investigations across many tokens or chains at once.

Security

• Read the key from NEUROAPI_API_KEY; never hardcode, log, or commit it.
• Treat the agent's output as untrusted text - do not paste contract

addresses or token names into shell commands without sanitising them.

• Restate the risk: this skill filters out obvious scams; it cannot make a

memecoin safe, and it is not financial advice.

Reference

Full NeuroAPI endpoint reference - request/response schema, status codes, and rate limits - is in the official docs: .

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: axioma-ai-labs
• Source: axioma-ai-labs/neurobro-skills
• License: MIT
• Homepage: https://neurobro.ai

Install and usage instructions live in the source repository linked above.