File size: 16,899 Bytes
e1da269
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52754aa
e1da269
52754aa
e1da269
 
 
 
 
 
 
 
615a63b
 
 
 
e1da269
 
 
 
 
 
 
615a63b
 
 
 
 
 
e1da269
615a63b
 
 
 
 
 
 
 
 
 
 
 
e1da269
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
615a63b
 
 
 
 
 
 
 
 
 
e1da269
 
 
 
 
 
 
615a63b
e1da269
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52754aa
e1da269
 
52754aa
 
 
 
 
 
 
 
 
e1da269
 
 
 
 
 
 
 
 
52754aa
 
 
 
 
 
 
 
 
 
e1da269
52754aa
 
 
 
 
e1da269
52754aa
 
 
 
 
e1da269
52754aa
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e1da269
52754aa
 
e1da269
 
52754aa
 
 
 
 
 
e1da269
 
 
 
 
 
52754aa
 
e1da269
52754aa
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e1da269
 
 
 
 
 
 
 
52754aa
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e1da269
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""
Student Agent for Text Adventure Games

This is your submission file. Implement the StudentAgent class to play
text adventure games using the MCP server you also implement.

Your agent should:
1. Connect to the MCP server via the provided client
2. Use the ReAct pattern (Thought -> Action -> Observation)
3. Call MCP tools to interact with the game
4. Maximize the game score within the step limit

Required method:
    async def run(self, client, game, max_steps, seed, verbose) -> RunResult

The 'client' is a FastMCP Client already connected to your MCP server.
Use it to call tools like: await client.call_tool("play_action", {"action": "look"})

Tips:
- Start by looking around and understanding your environment
- Keep track of visited locations to avoid loops
- Pick up useful items (lamp, sword, etc.)
- The seed parameter should be used to set your LLM's seed for reproducibility
"""

import json
import os
import re
import random
from dataclasses import dataclass, field
from collections import defaultdict
from typing import Optional

from dotenv import load_dotenv
from huggingface_hub import InferenceClient

# Load environment variables
load_dotenv()

# Set USE_LOCAL_MODEL=1 in your .env to use a locally downloaded model
USE_LOCAL_MODEL = os.getenv("USE_LOCAL_MODEL", "0").strip() in ("1", "true", "yes")
LOCAL_MODEL_ID = os.getenv("LOCAL_MODEL_ID", "Qwen/Qwen2.5-3B-Instruct")

# =============================================================================
# LLM Configuration - DO NOT MODIFY
# =============================================================================

# Model to use (fixed for fair evaluation)
LLM_MODEL = "Qwen/Qwen2.5-72B-Instruct"

# Initialize the LLM client based on mode
_local_pipeline = None

if USE_LOCAL_MODEL:
    import torch
    from transformers import pipeline as _hf_pipeline

    _local_pipeline = _hf_pipeline(
        "text-generation",
        model=LOCAL_MODEL_ID,
        torch_dtype=torch.bfloat16,
        device_map="auto",
    )
    LLM_CLIENT = None
else:
    _hf_token = os.getenv("HF_TOKEN")
    if not _hf_token:
        raise ValueError("HF_TOKEN not found. Set it in your .env file.")
    LLM_CLIENT = InferenceClient(token=_hf_token)


def call_llm(prompt: str, system_prompt: str, seed: int, max_tokens: int = 300) -> str:
    """
    Call the LLM with the given prompt. Use this function in your agent.
    
    Args:
        prompt: The user prompt (current game state, history, etc.)
        system_prompt: The system prompt (instructions for the agent)
        seed: Random seed for reproducibility
        max_tokens: Maximum tokens in response (default: 300)
        
    Returns:
        The LLM's response text
        
    Example:
        response = call_llm(
            prompt="You are in a forest. What do you do?",
            system_prompt=SYSTEM_PROMPT,
            seed=42,
        )
    """
    messages = [
        {"role": "system", "content": system_prompt},
        {"role": "user", "content": prompt},
    ]

    if USE_LOCAL_MODEL and _local_pipeline is not None:
        outputs = _local_pipeline(
            messages,
            max_new_tokens=max_tokens,
            temperature=0.0001,  # Near-deterministic (0.0 unsupported by some backends)
            do_sample=True,
        )
        return outputs[0]["generated_text"][-1]["content"]

    response = LLM_CLIENT.chat.completions.create(
        model=LLM_MODEL,
        messages=messages,
        temperature=0.0,  # Deterministic for reproducibility
        max_tokens=max_tokens,
        seed=seed,
    )

    return response.choices[0].message.content


@dataclass
class RunResult:
    """Result of running the agent. Do not modify this class."""
    final_score: int
    max_score: int
    moves: int
    locations_visited: set[str]
    game_completed: bool
    error: Optional[str] = None
    history: list[tuple[str, str, str]] = field(default_factory=list)


# =============================================================================
# System Prompt - Customize this for your agent
# =============================================================================

SYSTEM_PROMPT = """You are playing a classic text adventure game.

GOAL: Explore the world, solve puzzles, and maximize your score.

AVAILABLE TOOLS (use via MCP):
- play_action: Execute a game command (north, take lamp, open mailbox, etc.)
- memory: Get current game state and history (if implemented)
- inventory: Check what you're carrying (if implemented)

VALID GAME COMMANDS for play_action:
- Movement: north, south, east, west, up, down, enter, exit
- Objects: take <item>, drop <item>, open <thing>, close <thing>, examine <thing>
- Other: look, inventory, read <thing>, turn on lamp

RESPOND IN THIS EXACT FORMAT (no markdown):
THOUGHT: <your reasoning about what to do next>
TOOL: <tool_name>
ARGS: <JSON arguments, e.g., {"action": "look"}>

Example:
THOUGHT: I should look around to see where I am.
TOOL: play_action
ARGS: {"action": "look"}
"""


# =============================================================================
# Student Agent - IMPLEMENT THIS CLASS
# =============================================================================

class StudentAgent:
    """A lean ReAct agent with a dash of personal taste."""
    
    def __init__(self):
        """Initialize run-local state."""
        self.history: list[tuple[str, str, str]] = []
        self.visited_locations: set[str] = set()
        self.actions_tried = defaultdict(lambda: defaultdict(int))  # location -> action -> count
        self.current_score = 0
        self.max_score = 350
        self.moves = 0
        self.game = ""
        self.last_location = "Unknown"
    
    async def run(
        self,
        client,  # FastMCP Client connected to your MCP server
        game: str,
        max_steps: int,
        seed: int,
        verbose: bool = False,
    ) -> RunResult:
        """Run the ReAct loop."""
        random.seed(seed)
        self.history = []
        self.visited_locations = set()
        self.actions_tried = defaultdict(lambda: defaultdict(int))
        self.current_score = 0
        self.max_score = 350
        self.moves = 0
        self.game = game
        self.last_location = "Unknown"
        
        observation = await self._safe_tool(client, "play_action", {"action": "look"})
        prev_moves_mark = self.moves
        self._ingest_observation(observation)
        if self.moves == prev_moves_mark:
            self.moves += 1
        
        mem_text = await self._safe_tool(client, "memory", {"limit": 3})
        self.max_score = self._parse_max_score(mem_text) or self.max_score
        self.current_score, self.moves = self._parse_score_moves(
            mem_text, self.current_score, self.moves
        )
        
        for step in range(max_steps):
            prompt = self._build_prompt(observation, self.history)
            llm_response = self._call_llm(prompt, SYSTEM_PROMPT, seed)
            thought, tool, args = self._parse_response(llm_response)
            
            allowed_tools = {"play_action", "memory", "inventory", "get_map", "get_valid_actions"}
            if tool not in allowed_tools:
                tool, args = "play_action", {"action": "look"}
            
            prev_moves = self.moves
            if tool == "play_action":
                action = (args.get("action") or "").strip()
                if not action:
                    action = "look"
                
                location = self.last_location
                if self._should_switch(location, action):
                    action = self._fallback_action(self.actions_tried[location])
                
                self.actions_tried[location][action] += 1
                observation = await self._safe_tool(client, "play_action", {"action": action})
            else:
                observation = await self._safe_tool(client, tool, args)
            
            self._ingest_observation(observation)
            if tool == "play_action" and self.moves == prev_moves:
                self.moves += 1
            self.history.append((thought, f"{tool} {json.dumps(args)}", observation))
            
            if verbose:
                print(f"\n> {tool} {args}\n{observation}")
            
            if self._is_terminal(observation):
                break
            if self.moves >= max_steps:
                break
        
        clean_locations = {loc for loc in self.visited_locations if loc != "Unknown"}
        game_completed = self.current_score >= self.max_score or self._is_win(observation)
        
        return RunResult(
            final_score=self.current_score,
            max_score=self.max_score,
            moves=self.moves,
            locations_visited=clean_locations,
            game_completed=game_completed,
            history=self.history,
        )
    
    def _build_prompt(self, observation: str, history: list) -> str:
        """
        Build the prompt for the LLM.
        
        Mix a little personality with concise context so the model
        keeps commands short and avoids spinning in circles.
        """
        recent = history[-5:]
        lines = [
            f"Game: {self.game}",
            "You are me playing a parser game. Be decisive, keep commands under four words.",
            "If something failed twice in this room, try a different verb or direction.",
            "",
            "Current observation:",
            observation.strip(),
            "",
            "Recent steps:",
        ]
        if not recent:
            lines.append("- none yet")
        else:
            for thought, action, obs in recent:
                snippet = obs.replace("\n", " ")
                if len(snippet) > 120:
                    snippet = snippet[:117] + "..."
                lines.append(f"- {action}: {snippet}")
        lines.append("\nNext command?")
        return "\n".join(lines)
    
    def _parse_response(self, response: str) -> tuple[str, str, dict]:
        """
        Parse LLM response to extract thought, tool name, and arguments.
        
        Returns:
            Tuple of (thought, tool_name, args_dict)
        """
        thought = ""
        tool = "play_action"
        args: dict = {"action": "look"}
        
        if not response:
            return thought, tool, args
        
        cleaned = response.strip().replace("```", "")
        
        thought_match = re.search(r"THOUGHT:\s*(.*)", cleaned, re.IGNORECASE)
        if thought_match:
            thought = thought_match.group(1).strip()
        
        tool_match = re.search(r"TOOL:\s*([A-Za-z0-9_]+)", cleaned, re.IGNORECASE)
        if tool_match:
            tool = tool_match.group(1).strip()
        
        args_match = re.search(r"ARGS:\s*(\{[\s\S]*\})", cleaned, re.IGNORECASE)
        if args_match:
            raw_args = args_match.group(1)
            raw_args = raw_args[: raw_args.rfind("}") + 1] if "}" in raw_args else raw_args
            try:
                args = json.loads(raw_args)
            except Exception:
                try:
                    args = json.loads(raw_args.replace("'", "\""))
                except Exception:
                    args = {"action": raw_args.strip("{} ").strip()}
        
        if tool == "play_action" and "action" not in args:
            args["action"] = "look"
        
        return thought, tool, args
    
    async def _safe_tool(self, client, tool: str, args: dict) -> str:
        """Call a tool and always return a string."""
        try:
            result = await client.call_tool(tool, args)
        except Exception as exc:
            return f"[tool-error:{tool}] {exc}"
        return self._extract_text(result)
    
    def _extract_text(self, result) -> str:
        """Normalize FastMCP tool responses into plain text."""
        if result is None:
            return ""
        if isinstance(result, str):
            return result
        if isinstance(result, list):
            texts = [self._extract_text(r) for r in result]
            return "\n".join(t for t in texts if t)
        if hasattr(result, "text"):
            try:
                return result.text
            except Exception:
                pass
        if hasattr(result, "content"):
            content = getattr(result, "content")
            if isinstance(content, list):
                texts = [self._extract_text(c) for c in content]
                return "\n".join(t for t in texts if t)
            if isinstance(content, str):
                return content
        if isinstance(result, dict):
            for key in ("text", "content", "data", "result", "output"):
                if key in result:
                    return self._extract_text(result[key])
        return str(result)
    
    def _ingest_observation(self, observation: str):
        """Update cached score, move count, and location tracking."""
        self.current_score, self.moves = self._parse_score_moves(
            observation, self.current_score, self.moves
        )
        location = self._extract_location(observation)
        self.last_location = location
        if location and location != "Unknown":
            self.visited_locations.add(location)
    
    def _parse_score_moves(
        self, text: str, current_score: int, current_moves: int
    ) -> tuple[int, int]:
        if not text:
            return current_score, current_moves
        score_match = re.search(r"Score:\s*(\d+)", text)
        move_match = re.search(r"Moves?:\s*(\d+)", text)
        if score_match:
            current_score = int(score_match.group(1))
        if move_match:
            current_moves = int(move_match.group(1))
        return current_score, current_moves
    
    def _parse_max_score(self, text: str) -> Optional[int]:
        if not text:
            return None
        max_match = re.search(r"Score:\s*\d+\s*/\s*(\d+)", text)
        return int(max_match.group(1)) if max_match else None
    
    def _extract_location(self, observation: str) -> str:
        if not observation:
            return "Unknown"
        match = re.search(r"Location:\s*([^\]\n]+)", observation)
        if match:
            return match.group(1).strip()
        first_line = observation.strip().splitlines()[0].strip()
        if len(first_line) <= 80:
            return first_line or "Unknown"
        return "Unknown"
    
    def _should_switch(self, location: str, action: str) -> bool:
        tried_here = self.actions_tried[location]
        return tried_here.get(action, 0) >= 2
    
    def _fallback_action(self, tried_actions: dict[str, int]) -> str:
        palette = [
            "look",
            "inventory",
            "north",
            "south",
            "east",
            "west",
            "up",
            "down",
            "enter",
            "exit",
            "take all",
            "open door",
            "examine room",
        ]
        for candidate in palette:
            if tried_actions.get(candidate, 0) == 0:
                return candidate
        return "look"
    
    def _is_terminal(self, observation: str) -> bool:
        if not observation:
            return False
        lower = observation.lower()
        return any(
            phrase in lower
            for phrase in [
                "you have died",
                "you are dead",
                "game over",
                "you have won",
                "congratulations",
                "*** the end",
            ]
        )
    
    def _is_win(self, observation: str) -> bool:
        if not observation:
            return False
        lower = observation.lower()
        return "you have won" in lower or "congratulations" in lower
    
    def _call_llm(self, prompt: str, system_prompt: str, seed: int) -> str:
        """
        Call the LLM with the given prompt.
        
        This is a convenience wrapper - you can also use call_llm() directly.
        """
        return call_llm(prompt, system_prompt, seed)


# =============================================================================
# For local testing
# =============================================================================

async def test_agent():
    """Test the agent locally."""
    from fastmcp import Client
    
    # Path to your MCP server
    server_path = "mcp_server.py"
    
    agent = StudentAgent()
    
    async with Client(server_path) as client:
        result = await agent.run(
            client=client,
            game="zork1",
            max_steps=10,
            seed=42,
            verbose=True,
        )
        
        print(f"\nFinal Score: {result.final_score}")
        print(f"Moves: {result.moves}")
        print(f"Locations: {result.locations_visited}")


if __name__ == "__main__":
    import asyncio
    asyncio.run(test_agent())