Spaces:
Running
Running
File size: 20,690 Bytes
d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 a64025f d7caac8 a64025f d7caac8 a64025f d7caac8 b2d0640 d7caac8 b2d0640 a7b1897 adf245d b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 adf245d b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 a64025f d7caac8 a64025f d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 a7b1897 b2d0640 adf245d d7caac8 a64025f 5d60eec a64025f 5d60eec a64025f d7caac8 b2d0640 d7caac8 b2d0640 d484eff b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 a64025f b2d0640 a64025f d7caac8 b2d0640 a7b1897 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 a7b1897 b2d0640 a7b1897 b2d0640 a7b1897 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d484eff d7caac8 d484eff b2d0640 d7caac8 b2d0640 d484eff b2d0640 d484eff d7caac8 a64025f 5d60eec a64025f 5d60eec a64025f d7caac8 b2d0640 d7caac8 a64025f d7caac8 b2d0640 d7caac8 b2d0640 4910a59 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 5a6e59a 07fb7c8 b2d0640 c299edd 4910a59 c299edd b2d0640 d7caac8 7d0fa43 b2d9591 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 a64025f d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 b2d0640 d7caac8 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 | """
NyayaSetu V2 Agent — Full Intelligence Layer.
Pass 1 — ANALYSE: Understands message, detects tone/stage,
builds structured fact web, updates hypotheses,
forms targeted search queries, compresses summary.
Pass 2 — RETRIEVE: Parallel FAISS search. No LLM call.
Pass 3 — RESPOND: Dynamically assembled prompt + retrieved
context + full case state. Format-intelligent output.
2 LLM calls per turn. src/agent.py untouched.
"""
import os, sys, json, time, logging
from concurrent.futures import ThreadPoolExecutor, as_completed
from typing import Dict, Any, List
# sys.path must be set before any local imports
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from src.embed import embed_text
from src.retrieval import retrieve
from src.verify import verify_citations
from src.system_prompt import build_prompt, ANALYSIS_PROMPT
from src.ner import extract_entities, augment_query
logger = logging.getLogger(__name__)
from tenacity import retry, stop_after_attempt, wait_exponential, retry_if_exception_type
from dotenv import load_dotenv
import threading
import time
from src.llm import call_llm_raw
load_dotenv()
# ── Circuit Breaker for Groq API ──────────────────────────
class CircuitBreaker:
"""Simple circuit breaker to detect when Groq API is down."""
def __init__(self, failure_threshold=5, recovery_timeout=60):
self.failure_count = 0
self.failure_threshold = failure_threshold
self.recovery_timeout = recovery_timeout
self.last_failure_time = None
self.is_open = False
self.lock = threading.Lock()
def record_success(self):
with self.lock:
self.failure_count = 0
self.is_open = False
def record_failure(self):
with self.lock:
self.failure_count += 1
self.last_failure_time = time.time()
if self.failure_count >= self.failure_threshold:
self.is_open = True
logger.warning(f"Circuit breaker OPEN: {self.failure_count} failures detected")
def can_attempt(self) -> bool:
with self.lock:
if not self.is_open:
return True
# Try to recover after timeout
if time.time() - self.last_failure_time > self.recovery_timeout:
logger.info("Circuit breaker attempting recovery...")
self.is_open = False
self.failure_count = 0
return True
return False
def get_status(self) -> str:
with self.lock:
if self.is_open:
return f"OPEN ({self.failure_count} failures)"
return f"CLOSED ({self.failure_count} failures)"
_circuit_breaker = CircuitBreaker()
# ── Session store ─────────────────────────────────────────
sessions: Dict[str, Dict] = {}
def empty_case_state() -> Dict:
return {
"parties": [],
"events": [],
"documents": [],
"amounts": [],
"locations": [],
"timeline": [],
"disputes": [],
"hypotheses": [],
"stage": "intake",
"last_response_type": "none",
"turn_count": 0,
"facts_missing": [],
"context_interpreted": False,
"last_radar_turn": -3, # track when radar last fired
"last_format": "none",
"format_override_turn": -1, # track when user explicitly requested a format
}
def get_or_create_session(session_id: str) -> Dict:
if session_id not in sessions:
sessions[session_id] = {
"summary": "",
"last_3_messages": [],
"case_state": empty_case_state()
}
return sessions[session_id]
def update_session(session_id: str, analysis: Dict, user_message: str, response: str):
session = sessions[session_id]
cs = session["case_state"]
if analysis.get("updated_summary"):
session["summary"] = analysis["updated_summary"]
facts = analysis.get("facts_extracted", {})
if facts:
for key in ["parties", "events", "documents", "amounts", "locations", "disputes"]:
new_items = facts.get(key, [])
existing = cs.get(key, [])
for item in new_items:
if item and item not in existing:
existing.append(item)
cs[key] = existing
for ev in facts.get("timeline_events", []):
if ev and ev not in cs["timeline"]:
cs["timeline"].append(ev)
for nh in analysis.get("hypotheses", []):
existing_claims = [h["claim"] for h in cs["hypotheses"]]
if nh.get("claim") and nh["claim"] not in existing_claims:
cs["hypotheses"].append(nh)
else:
for h in cs["hypotheses"]:
if h["claim"] == nh.get("claim"):
h["confidence"] = nh.get("confidence", h["confidence"])
for e in nh.get("evidence", []):
if e not in h.get("evidence", []):
h.setdefault("evidence", []).append(e)
cs["stage"] = analysis.get("stage", cs["stage"])
cs["last_response_type"] = analysis.get("action_needed", "none")
cs["facts_missing"] = analysis.get("facts_missing", [])
cs["last_format"] = analysis.get("format_decision", "none")
cs["turn_count"] = cs.get("turn_count", 0) + 1
if cs["turn_count"] >= 3:
cs["context_interpreted"] = True
session["last_3_messages"].append({"role": "user", "content": user_message})
session["last_3_messages"].append({"role": "assistant", "content": response[:400]})
if len(session["last_3_messages"]) > 6:
session["last_3_messages"] = session["last_3_messages"][-6:]
# ── Pass 1: Analyse ───────────────────────────────────────
# Retry up to 5 times with exponential backoff (1s to 16s) to handle transient failures
@retry(stop=stop_after_attempt(5), wait=wait_exponential(min=1, max=16, multiplier=1.5))
def analyse(user_message: str, session: Dict) -> Dict:
if not _circuit_breaker.can_attempt():
logger.error(f"Circuit breaker OPEN - skipping Pass 1. Status: {_circuit_breaker.get_status()}")
raise Exception("Groq API circuit breaker is open - service unavailable")
summary = session.get("summary", "")
last_msgs = session.get("last_3_messages", [])
cs = session["case_state"]
last_response_type = cs.get("last_response_type", "none")
turn_count = cs.get("turn_count", 0)
history_text = "\n".join(
f"{m['role'].upper()}: {m['content'][:250]}"
for m in last_msgs[-4:]
) if last_msgs else ""
fact_web = ""
if any(cs.get(k) for k in ["parties", "events", "documents", "amounts", "disputes"]):
hyp_lines = "\n".join(
f" - {h['claim']} [{h.get('confidence','?')}]"
for h in cs.get("hypotheses", [])[:3]
) or " none yet"
fact_web = f"""
CURRENT FACT WEB:
- Parties: {', '.join(cs.get('parties', [])) or 'none'}
- Events: {', '.join(cs.get('events', [])) or 'none'}
- Documents/Evidence: {', '.join(cs.get('documents', [])) or 'none'}
- Amounts: {', '.join(cs.get('amounts', [])) or 'none'}
- Disputes: {', '.join(cs.get('disputes', [])) or 'none'}
- Active hypotheses:
{hyp_lines}"""
user_content = f"""CONVERSATION SUMMARY:
{summary if summary else "First message — no prior context."}
RECENT MESSAGES:
{history_text if history_text else "None"}
LAST RESPONSE TYPE: {last_response_type}
TURN COUNT: {turn_count}
{fact_web}
NEW USER MESSAGE:
{user_message}
Rules:
- If last_response_type was "question", action_needed CANNOT be "question"
- action_needed SHOULD differ from last_response_type for variety
- Extract ALL facts from user message even if implied
- Update hypothesis confidence based on new evidence
- search_queries must be specific legal questions for vector search
- format_decision must be chosen fresh each turn based on THIS message's content
- NEVER carry over format_decision from previous turn unless user explicitly requests it again
- If user requested a specific format last turn, revert to most natural format this turn"""
response = call_llm_raw(
messages=[
{"role": "system", "content": ANALYSIS_PROMPT},
{"role": "user", "content": user_content}
]
)
_circuit_breaker.record_success() # API call succeeded
raw = response.strip()
raw = raw.replace("```json", "").replace("```", "").strip()
try:
analysis = json.loads(raw)
except json.JSONDecodeError:
logger.warning(f"Pass 1 JSON parse failed: {raw[:200]}")
analysis = {
"tone": "casual", "format_requested": "none",
"subject": "legal query", "action_needed": "advice",
"urgency": "medium",
"hypotheses": [{"claim": user_message[:80], "confidence": "low", "evidence": []}],
"facts_extracted": {}, "facts_missing": [],
"legal_issues": [], "clarifying_question": {},
"stage": "understanding", "last_response_type": last_response_type,
"updated_summary": f"{summary} | {user_message[:100]}",
"search_queries": [user_message[:200]],
"should_interpret_context": False,
"format_decision": "none"
}
return analysis
# ── Pass 2: Retrieve ──────────────────────────────────────
def retrieve_parallel(search_queries: List[str], top_k: int = 5) -> List[Dict]:
if not search_queries:
return []
all_results = []
def search_one(query):
try:
embedding = embed_text(query)
return retrieve(embedding, top_k=top_k)
except Exception as e:
logger.warning(f"FAISS search failed: {e}")
return []
with ThreadPoolExecutor(max_workers=min(3, len(search_queries))) as executor:
futures = {executor.submit(search_one, q): q for q in search_queries}
for future in as_completed(futures):
all_results.extend(future.result())
seen = {}
for chunk in all_results:
cid = chunk.get("chunk_id") or chunk.get("judgment_id", "")
score = chunk.get("similarity_score", 999)
if cid not in seen or score < seen[cid]["similarity_score"]:
seen[cid] = chunk
return sorted(seen.values(), key=lambda x: x.get("similarity_score", 999))[:top_k]
# ── Pass 3: Respond ───────────────────────────────────────
# Retry up to 5 times with exponential backoff (2s to 32s) — more aggressive than Pass 1
@retry(stop=stop_after_attempt(5), wait=wait_exponential(min=2, max=32, multiplier=1.5))
def respond(user_message: str, analysis: Dict, chunks: List[Dict], session: Dict) -> str:
if not _circuit_breaker.can_attempt():
logger.error(f"Circuit breaker OPEN - skipping Pass 3. Status: {_circuit_breaker.get_status()}")
raise Exception("Groq API circuit breaker is open - service unavailable")
system_prompt = build_prompt(analysis)
cs = session["case_state"]
turn_count = cs.get("turn_count", 0)
context_parts = []
for chunk in chunks[:5]:
source_type = chunk.get("source_type", "case_law")
title = chunk.get("title", "Unknown")
year = chunk.get("year", "")
jid = chunk.get("judgment_id", "")
text = chunk.get("expanded_context") or chunk.get("chunk_text") or chunk.get("text", "")
type_labels = {
"statute": f"[STATUTE: {title} | {year}]",
"procedure": f"[PROCEDURE: {title}]",
"law_commission": f"[LAW COMMISSION: {title}]",
"legal_reference": f"[LEGAL REFERENCE: {title}]",
"statute_qa": f"[LEGAL QA: {title}]",
}
header = type_labels.get(source_type, f"[CASE: {title} | {year} | {jid}]")
context_parts.append(f"{header}\n{text[:800]}")
context = "\n\n".join(context_parts) if context_parts else "No relevant sources retrieved."
case_summary = ""
if cs.get("parties") or cs.get("hypotheses"):
hyp_text = "\n".join(
f" - {h['claim']} [{h.get('confidence','?')} confidence] "
f"| evidence: {', '.join(h.get('evidence', [])) or 'none yet'}"
for h in cs.get("hypotheses", [])[:4]
) or " none established"
case_summary = f"""
CASE STATE (built across {turn_count} turns):
Parties: {', '.join(cs.get('parties', [])) or 'unspecified'}
Events: {', '.join(cs.get('events', [])) or 'unspecified'}
Evidence: {', '.join(cs.get('documents', [])) or 'none mentioned'}
Amounts: {', '.join(cs.get('amounts', [])) or 'none'}
Active hypotheses:
{hyp_text}
Missing facts: {', '.join(cs.get('facts_missing', [])) or 'none critical'}
Stage: {cs.get('stage', 'intake')}"""
# Context interpretation — only once per conversation at turn 2
interpret_instruction = ""
should_interpret = analysis.get("should_interpret_context", False)
if should_interpret and not cs.get("context_interpreted") and turn_count == 2:
interpret_instruction = "\nIn one sentence only, reflect back your understanding of the situation before responding."
# Radar — only fires every 3 turns, not every turn
last_radar_turn = cs.get("last_radar_turn", -3)
if (turn_count - last_radar_turn) >= 3:
cs["last_radar_turn"] = turn_count
radar_instruction = """
PROACTIVE RADAR — only if a genuinely non-obvious legal angle exists that hasn't been mentioned yet:
Add a single "⚡ You Should Also Know:" line (1-2 sentences max).
Skip entirely if the response already covers all relevant angles or if this is a question/understanding turn."""
else:
radar_instruction = "Do NOT add a 'You Should Also Know' section this turn."
summary = session.get("summary", "")
last_msgs = session.get("last_3_messages", [])
history_text = "\n".join(
f"{m['role'].upper()}: {m['content'][:300]}"
for m in last_msgs[-4:]
) if last_msgs else ""
user_content = f"""CONVERSATION SUMMARY:
{summary if summary else "First message."}
RECENT CONVERSATION:
{history_text if history_text else "None"}
{case_summary}
RETRIEVED LEGAL SOURCES:
{context}
USER MESSAGE: {user_message}
THIS TURN:
- Legal hypotheses: {', '.join(h['claim'] for h in analysis.get('hypotheses', [])[:3]) or 'analysing'}
- Stage: {analysis.get('stage', 'understanding')}
- Urgency: {analysis.get('urgency', 'medium')}
- Response type: {analysis.get('action_needed', 'advice')}
- Format: {analysis.get('format_decision', 'appropriate for content')}
{interpret_instruction}
Instructions:
- Cite specific sources when making legal claims
- Use your legal knowledge for reasoning and context
- Format: {analysis.get('format_decision', 'use the most appropriate format for the content type')}
- Opposition war-gaming: if giving strategy, include what the other side will argue
{radar_instruction}"""
response = call_llm_raw(
messages=[
{"role": "system", "content": system_prompt},
{"role": "user", "content": user_content}
]
)
_circuit_breaker.record_success() # API call succeeded
return response
# ── Main entry point ──────────────────────────────────────
def run_query_v2(user_message: str, session_id: str) -> Dict[str, Any]:
start = time.time()
session = get_or_create_session(session_id)
# Pass 1
try:
analysis = analyse(user_message, session)
except Exception as e:
error_type = type(e).__name__
logger.error(f"Pass 1 failed after retries: {error_type}: {e}. Circuit breaker: {_circuit_breaker.get_status()}")
# Record API failure if it was a connection error
if "APIConnectionError" in error_type or "RateLimitError" in error_type:
_circuit_breaker.record_failure()
analysis = {
"tone": "casual", "format_requested": "none",
"subject": "legal query", "action_needed": "advice",
"urgency": "medium",
"hypotheses": [{"claim": user_message[:80], "confidence": "low", "evidence": []}],
"facts_extracted": {}, "facts_missing": [],
"legal_issues": [], "clarifying_question": {},
"stage": "understanding", "last_response_type": "none",
"updated_summary": user_message[:200],
"search_queries": [user_message[:200]],
"should_interpret_context": False,
"format_decision": "none"
}
# Extract entities and augment queries for better retrieval
entities = extract_entities(user_message)
augmented_message = augment_query(user_message, entities)
# Pass 2 — build search queries from analysis + legal issues
search_queries = analysis.get("search_queries", [augmented_message])
if not search_queries:
search_queries = [augmented_message]
# Add queries from issue spotter
for issue in analysis.get("legal_issues", []):
statutes = issue.get("relevant_statutes", [])
specific = issue.get("specific_issue", "")
if specific:
issue_query = f"{specific} {' '.join(statutes[:2])}".strip()
if issue_query not in search_queries:
search_queries.append(issue_query)
if augmented_message not in search_queries:
search_queries.append(augmented_message)
chunks = []
try:
# Retrieve more candidates for reranker to work with
raw_chunks = retrieve_parallel(search_queries[:3], top_k=10)
# Rerank candidates by true relevance
from src.reranker import rerank
chunks = rerank(user_message, raw_chunks, top_k=5)
# Add precedent chain
from src.citation_graph import get_precedent_chain
retrieved_ids = [c.get("judgment_id", "") for c in chunks]
precedents = get_precedent_chain(retrieved_ids, max_precedents=2)
if precedents:
chunks.extend(precedents)
except Exception as e:
logger.error(f"Pass 2 failed: {e}")
# Pass 3
try:
answer = respond(user_message, analysis, chunks, session)
except Exception as e:
error_type = type(e).__name__
logger.error(f"Pass 3 failed after retries: {error_type}: {e}. Circuit breaker: {_circuit_breaker.get_status()}")
# Record API failure if it was a connection error
if "APIConnectionError" in error_type or "RateLimitError" in error_type:
_circuit_breaker.record_failure()
if chunks:
fallback = "\n\n".join(
f"[{c.get('title', 'Source')}]\n{c.get('text', '')[:400]}"
for c in chunks[:3]
)
answer = f"LLM service temporarily unavailable. Most relevant excerpts:\n\n{fallback}"
else:
answer = "I encountered an issue processing your request. Please try again."
verification_status, unverified_quotes = verify_citations(answer, chunks)
update_session(session_id, analysis, user_message, answer)
sources = []
for c in chunks:
title = c.get("title", "")
jid = c.get("judgment_id", "")
sources.append({
"meta": {
"judgment_id": jid,
"title": title if title and title != jid else jid,
"year": c.get("year", ""),
"chunk_index": c.get("chunk_index", 0),
"source_type": c.get("source_type", "case_law"),
"court": c.get("court", "Supreme Court of India")
},
"text": (c.get("expanded_context") or c.get("chunk_text") or c.get("text", ""))[:600]
})
return {
"query": user_message,
"answer": answer,
"sources": sources,
"verification_status": verification_status,
"unverified_quotes": unverified_quotes,
"entities": entities,
"num_sources": len(chunks),
"truncated": False,
"session_id": session_id,
"analysis": {
"tone": analysis.get("tone"),
"stage": analysis.get("stage"),
"urgency": analysis.get("urgency"),
"hypotheses": [h["claim"] for h in analysis.get("hypotheses", [])]
}
} |