config.toml

# Telegram Channel Crawler Configuration

[crawler]
# How many days back in time to look for messages
time_window_days = 365

# Maximum number of messages to fetch per channel
max_messages_per_channel = 40000

# Number of channels to process concurrently
parallel_requests = 1

# Number of messages to fetch per batch
batch_size = 500

# Delay between requests in seconds (helps avoid rate limiting)
rate_limiting_delay = 0.5

# Save checkpoint every N messages (0 to disable checkpoints)
# Checkpoints allow resuming if the script is interrupted
checkpoint_interval = 2000

# Fetch replies/comments to channel posts (true/false)
# For channels with discussion groups, this fetches all replies to posts
fetch_replies = true

# Maximum depth for nested replies (0 = no replies, 1 = direct replies only, 2 = replies to replies, etc.)
# Higher values fetch more nested conversations but take longer
max_reply_depth = 4

[group_chats]
# List of group chat to crawl (must be numeric IDs, not usernames)
# Use 'python list_channels.py' to see all your channels and their IDs
include = [
    2244967395, # Base LatAm
    1242127973, # Lobster DAO
    1533865579, # Decentralised.co
]

[channels]
# List of channel IDs to crawl (must be numeric IDs, not usernames)
# Use 'python list_channels.py' to see all your channels and their IDs
include = [
    2468695644, # Ritual-Korea - Channel
]

[output]
# Pretty print JSON output (uses indentation)
pretty_print = true

# JSON indentation spaces (only used if pretty_print is true)
indent_spaces = 2

# Note: Messages are saved to raw/[channel_id]_messages.json

[trust]
# Trust scoring values
mention_points = 50      # Points for direct mentions
reply_points = 40        # Points for replies
reaction_points = 30     # Points for reactions