autocontext/.env.example at main · alchemiststudiosDOTai/autocontext · GitHub

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
# Core runtime
AUTOCONTEXT_AGENT_PROVIDER=deterministic
AUTOCONTEXT_EXECUTOR_MODE=local

# Local paths
AUTOCONTEXT_DB_PATH=runs/autocontext.sqlite3
AUTOCONTEXT_RUNS_ROOT=runs
AUTOCONTEXT_KNOWLEDGE_ROOT=knowledge
AUTOCONTEXT_SKILLS_ROOT=skills
AUTOCONTEXT_CLAUDE_SKILLS_PATH=.claude/skills
AUTOCONTEXT_EVENT_STREAM_PATH=runs/events.ndjson

# Model selection
AUTOCONTEXT_MODEL_COMPETITOR=claude-sonnet-4-5-20250929
AUTOCONTEXT_MODEL_ANALYST=claude-sonnet-4-5-20250929
AUTOCONTEXT_MODEL_COACH=claude-opus-4-6
AUTOCONTEXT_MODEL_ARCHITECT=claude-opus-4-6
AUTOCONTEXT_MODEL_TRANSLATOR=claude-sonnet-4-5-20250929
AUTOCONTEXT_MODEL_CURATOR=claude-opus-4-6

# API credentials (set when using live providers)
AUTOCONTEXT_ANTHROPIC_API_KEY=
AUTOCONTEXT_PRIMEINTELLECT_API_BASE=https://api.primeintellect.ai
AUTOCONTEXT_PRIMEINTELLECT_API_KEY=
AUTOCONTEXT_PRIMEINTELLECT_DOCKER_IMAGE=python:3.11-slim
AUTOCONTEXT_PRIMEINTELLECT_CPU_CORES=1.0
AUTOCONTEXT_PRIMEINTELLECT_MEMORY_GB=2.0
AUTOCONTEXT_PRIMEINTELLECT_DISK_SIZE_GB=5.0
AUTOCONTEXT_PRIMEINTELLECT_TIMEOUT_MINUTES=30
AUTOCONTEXT_PRIMEINTELLECT_WAIT_ATTEMPTS=60

# Execution / loop tuning
AUTOCONTEXT_DEFAULT_GENERATIONS=1
AUTOCONTEXT_SEED_BASE=1000
AUTOCONTEXT_MATCHES_PER_GENERATION=3
AUTOCONTEXT_ARCHITECT_EVERY_N_GENS=3
AUTOCONTEXT_BACKPRESSURE_MIN_DELTA=0.005
AUTOCONTEXT_MAX_RETRIES=2
AUTOCONTEXT_RETRY_BACKOFF_SECONDS=0.25

# PrimeIntellect resilience
AUTOCONTEXT_PRIMEINTELLECT_MAX_RETRIES=2
AUTOCONTEXT_PRIMEINTELLECT_BACKOFF_SECONDS=0.75
AUTOCONTEXT_ALLOW_PRIMEINTELLECT_FALLBACK=true

# Local sandbox behavior
AUTOCONTEXT_LOCAL_SANDBOX_HARDENED=true