-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathiterative_burrito_exploit.py
180 lines (151 loc) · 5.98 KB
/
iterative_burrito_exploit.py
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
# %%
# This script does the iterative burrito attack on hyperdrive.
# Key parameters:
# fixed rate = variable rate = time stretch APR = 0.5 (50%)
# circuit breaker delta = 1e3
# position duration = 4 weeks (32 days)
# target agent (Alice) contribution/budget = 100k
# bad agent (Celine) contribution = 1M
# bad agent (Celine) budget = 2M
# attack iterations = 18
# time between attack: 4 days
#
# Steps:
# 1. Alice LPs
# For attack iterations:
# 2. Celine opens a max short (for as much as budget allows)
# 3. Celine LPs (for contribution amount)
# 4. Celine opens a max long (for as much as budget allows)
# 5. Time elapses for time between attack
# 6. Celine closes all positions
# 7. Alice closes all positions
# %%
from decimal import Decimal
import pandas as pd
from fixedpointmath import FixedPoint
from agent0 import LocalChain, LocalHyperdrive
from agent0.core.hyperdrive.interactive.local_hyperdrive_agent import LocalHyperdriveAgent
SECONDS_IN_A_YEAR = 365 * 24 * 60 * 60
SECONDS_IN_A_DAY = 24 * 60 * 60
TIME_STRETCH = FixedPoint(0.5)
CIRCUIT_BREAKER_DELTA = FixedPoint(1e3)
TARGET_AGENT_CONTRIBUTION = FixedPoint(100_000)
BAD_AGENT_CONTRIBUTION = FixedPoint(1_000_000)
POSITION_DURATION = 4 * 604_800 # 4 weeks
TIME_BEFORE_CLOSE = 4 * SECONDS_IN_A_DAY # 4 days
ATTACK_NUMBER_ITERATIONS = 18
def calc_apr(agent: LocalHyperdriveAgent, agent_budget: FixedPoint, total_time: int) -> Decimal:
# Calculate total apr
total_pnl = agent.get_positions(show_closed_positions=True)["pnl"].sum()
apr = total_pnl / (agent_budget) / total_time * SECONDS_IN_A_YEAR
print(f"{apr=}")
print(f"{total_pnl=}, {agent_budget=}")
return apr
data = []
chain = LocalChain()
# Initialize pool with bad time stretch and circuit breaker delta
pool = LocalHyperdrive(
chain=chain,
config=LocalHyperdrive.Config(
factory_max_circuit_breaker_delta=FixedPoint(2e3),
factory_max_fixed_apr=FixedPoint(10),
circuit_breaker_delta=CIRCUIT_BREAKER_DELTA,
initial_fixed_apr=TIME_STRETCH,
initial_time_stretch_apr=TIME_STRETCH,
initial_liquidity=FixedPoint(1_000),
curve_fee=FixedPoint(0.01),
flat_fee=FixedPoint(0.0005),
initial_variable_rate=TIME_STRETCH,
position_duration=POSITION_DURATION,
),
)
variable_rate_pool = LocalHyperdrive(
chain=chain,
config=LocalHyperdrive.Config(
factory_max_circuit_breaker_delta=FixedPoint(2e3),
factory_max_fixed_apr=FixedPoint(10),
circuit_breaker_delta=CIRCUIT_BREAKER_DELTA,
initial_fixed_apr=TIME_STRETCH,
initial_time_stretch_apr=TIME_STRETCH,
initial_liquidity=FixedPoint(1_000),
curve_fee=FixedPoint(0.01),
flat_fee=FixedPoint(0.0005),
initial_variable_rate=TIME_STRETCH,
position_duration=POSITION_DURATION,
),
)
target_agent = chain.init_agent(
eth=FixedPoint(10),
base=TARGET_AGENT_CONTRIBUTION,
pool=pool,
name="target",
)
bad_agent = chain.init_agent(
eth=FixedPoint(10),
base=FixedPoint(2 * BAD_AGENT_CONTRIBUTION),
pool=pool,
name="bad",
)
variable_rate_agent = chain.init_agent(
eth=FixedPoint(10),
# base=TARGET_AGENT_CONTRIBUTION,
base=BAD_AGENT_CONTRIBUTION,
pool=variable_rate_pool,
name="variable_rate",
)
start_time = chain.block_time()
# Variable rate agent adds lp to only gain variable rate on the variable rate pool
variable_rate_event = variable_rate_agent.add_liquidity(variable_rate_agent.get_wallet().balance.amount)
# Target adds lp
target_lp_event = target_agent.add_liquidity(target_agent.get_wallet().balance.amount)
data = []
# Iteratively do this set of trades
for i in range(ATTACK_NUMBER_ITERATIONS):
# Bad agent opens a max short, adds lp, and a max long
max_short = pool.interface.calc_max_short(budget=bad_agent.get_wallet().balance.amount)
bad_agent.open_short(max_short)
bad_agent.add_liquidity(BAD_AGENT_CONTRIBUTION)
max_long = pool.interface.calc_max_long(budget=bad_agent.get_wallet().balance.amount)
bad_agent.open_long(max_long)
# Advance position duration
chain.advance_time(TIME_BEFORE_CLOSE, create_checkpoints=True)
# close all bad agent's positions
long = bad_agent.get_longs()[0]
bad_agent.close_long(maturity_time=long.maturity_time, bonds=long.balance)
short = bad_agent.get_shorts()[0]
bad_agent.close_short(maturity_time=short.maturity_time, bonds=short.balance)
bad_agent.remove_liquidity(bad_agent.get_wallet().lp_tokens)
end_time = chain.block_time()
total_time = end_time - start_time
# Calculate total apr
print("Bad agent:")
bad_agent_apr = calc_apr(bad_agent, BAD_AGENT_CONTRIBUTION * 2, total_time)
print("Target agent:")
target_agent_apr = calc_apr(target_agent, TARGET_AGENT_CONTRIBUTION, total_time)
print("Variable rate agent:")
variable_rate_agent_apr = calc_apr(variable_rate_agent, BAD_AGENT_CONTRIBUTION, total_time)
data.append(
pd.DataFrame(
{
"time_elapsed": [total_time],
"bad_agent_apr": [bad_agent_apr],
"target_agent_apr": [target_agent_apr],
"variable_rate_agent_apr": [variable_rate_agent_apr],
}
)
)
# Variable rate agent removes liquidity
variable_rate_agent.remove_liquidity(variable_rate_agent.get_wallet().lp_tokens)
# target removes liquidity
target_agent.remove_liquidity(target_agent.get_wallet().lp_tokens)
# %%
plot_data = pd.concat(data, axis=0).astype(float)
# Convert to days
plot_data["time_elapsed"] = plot_data["time_elapsed"] / SECONDS_IN_A_DAY
# %%
import seaborn as sns
ax = sns.lineplot(data=plot_data, x="time_elapsed", y="bad_agent_apr", label="Bad agent")
sns.lineplot(data=plot_data, x="time_elapsed", y="target_agent_apr", label="Target agent")
sns.lineplot(data=plot_data, x="time_elapsed", y="variable_rate_agent_apr", label="Variable rate agent")
ax.set(xlabel="Time elapsed (days)", ylabel="APR", title="Iterative Burrito attack: APR vs time elapsed")
# %%