c762564d40
- Fix 71 invalid-syntax files (class-body newline-broken assignments) - Add from/None chain to 307 B904 raise-without-from sites - Add B008 ignore to ruff.toml (already in pyproject.toml) - Noqa F401 on __init__.py re-exports (137 sites) - Noqa E402 on deferred imports (63 sites) - Bulk-add stdlib/FastAPI/project imports for F821 (127 sites) - Replace ×→x, –→-, …→... in docstrings (4093 chars) - Manual refactor of 5 SIM103/SIM116 patterns Tests: 791 passed (66 deselected due to pre-existing Redis issues in test_rag.py) Co-authored-by: opencode <opencode@rugmunch.io>
113 lines
4.1 KiB
Python
113 lines
4.1 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
RMI AI Pipeline Part 2 - Remaining 7 Modules
|
|
=============================================
|
|
Community Forensics | Cross-Chain Entity | Ghost Blog | Social Media | Token Compare
|
|
All Ollama Cloud deepseek-v4-flash. ~$0.001/operation.
|
|
"""
|
|
|
|
import json
|
|
import logging
|
|
import os
|
|
from urllib.request import Request, urlopen
|
|
|
|
logger = logging.getLogger("rmi.ai_pipeline2")
|
|
OLLAMA_KEY = os.getenv("OLLAMA_API_KEY", os.getenv("DEEPSEEK_API_KEY", ""))
|
|
OLLAMA_URL = "https://ollama.com/v1/chat/completions"
|
|
MODEL = "deepseek-v4-flash"
|
|
|
|
|
|
def _call_ai(system: str, prompt: str, max_tokens: int = 250, temp: float = 0.3) -> str:
|
|
try:
|
|
body = json.dumps(
|
|
{
|
|
"model": MODEL,
|
|
"messages": [
|
|
{"role": "system", "content": system},
|
|
{"role": "user", "content": prompt},
|
|
],
|
|
"max_tokens": max_tokens,
|
|
"temperature": temp,
|
|
}
|
|
).encode()
|
|
req = Request(
|
|
OLLAMA_URL,
|
|
data=body,
|
|
headers={"Authorization": f"Bearer {OLLAMA_KEY}", "Content-Type": "application/json"},
|
|
)
|
|
resp = urlopen(req, timeout=15)
|
|
return json.loads(resp.read())["choices"][0]["message"]["content"].strip()
|
|
except Exception as e:
|
|
logger.error(f"AI call failed: {e}")
|
|
return ""
|
|
|
|
|
|
# ── 8. COMMUNITY FORENSICS AUTO-ANALYSIS ──
|
|
FORENSICS_SYSTEM = """You are a crypto forensics investigator. A community member submitted a suspicious token for review.
|
|
Analyze the information and provide:
|
|
1. Initial verdict (LIKELY SCAM / SUSPICIOUS / NEEDS MORE INFO)
|
|
2. Key concerns (2-3 bullet points)
|
|
3. Recommended next steps for the investigator
|
|
Keep it under 150 words."""
|
|
|
|
|
|
def analyze_community_submission(submission: dict) -> str:
|
|
return _call_ai(FORENSICS_SYSTEM, json.dumps(submission)[:1500], max_tokens=250)
|
|
|
|
|
|
# ── 10. CROSS-CHAIN ENTITY DETECTION ──
|
|
CROSSCHAIN_SYSTEM = """You identify crypto entities operating across multiple blockchains.
|
|
Given wallet data from different chains, determine if they're the same entity.
|
|
Reply format: MATCH|confidence_0-100|reason OR NO_MATCH|reason"""
|
|
|
|
|
|
def detect_cross_chain(wallets: dict) -> str:
|
|
return _call_ai(CROSSCHAIN_SYSTEM, json.dumps(wallets)[:1500], max_tokens=100)
|
|
|
|
|
|
# ── 11. GHOST BLOG AUTO-DRAFT ──
|
|
GHOST_SYSTEM = """You are a crypto security blogger for Rug Munch Intelligence (rugmunch.io).
|
|
Write a blog post draft from scanner data and incident reports.
|
|
Structure:
|
|
- Title (catchy, SEO-friendly, under 80 chars)
|
|
- Hook (1 sentence that grabs attention)
|
|
- Body (3-4 paragraphs explaining the threat)
|
|
- Key takeaways (2-3 bullet points)
|
|
- Call to action (check your tokens, use our scanner)
|
|
Use markdown formatting. Professional but engaging tone."""
|
|
|
|
|
|
def draft_blog_post(topic: str, data: dict) -> str:
|
|
prompt = f"Topic: {topic}\n\nData:\n{json.dumps(data)[:2000]}"
|
|
return _call_ai(GHOST_SYSTEM, prompt, max_tokens=500, temp=0.6)
|
|
|
|
|
|
# ── 13. SOCIAL MEDIA POST GENERATOR ──
|
|
SOCIAL_SYSTEM = """You are the social media manager for Rug Munch Intelligence (@CryptoRugMunch).
|
|
Write a tweet/telegram post about a crypto security finding.
|
|
Rules:
|
|
- Under 280 chars for Twitter, under 500 for Telegram
|
|
- Start with a hook (stat, warning, or question)
|
|
- Include $TICKER if relevant
|
|
- End with a call to action or link
|
|
- Use emojis sparingly (1-2 max)
|
|
- No hashtag spam (2-3 max)
|
|
Reply format: TWITTER: <tweet> | TELEGRAM: <post>"""
|
|
|
|
|
|
def generate_social_post(incident: dict, platform: str = "both") -> str:
|
|
return _call_ai(SOCIAL_SYSTEM, json.dumps(incident)[:1000], max_tokens=200, temp=0.7)
|
|
|
|
|
|
# ── 14. TOKEN COMPARISON ENGINE ──
|
|
COMPARE_SYSTEM = """Compare two crypto tokens for safety. Given their scanner results, determine which is safer and why.
|
|
Reply format:
|
|
SAFER: <token_name>
|
|
REASON: <2-3 sentence comparison>
|
|
SCORE_DIFF: <token1_score> vs <token2_score>
|
|
KEY_DIFFERENCES: <bullet points>"""
|
|
|
|
|
|
def compare_tokens(token_a: dict, token_b: dict) -> str:
|
|
prompt = f"Token A:\n{json.dumps(token_a)[:800]}\n\nToken B:\n{json.dumps(token_b)[:800]}"
|
|
return _call_ai(COMPARE_SYSTEM, prompt, max_tokens=200)
|