Struct TelegramConfig 

pub struct TelegramConfig {
    pub token: Option<String>,
    pub allowed_users: Vec<String>,
    pub skills: ChannelSkillsConfig,
    pub allowed_tools: Option<Vec<String>>,
    pub stream_interval_ms: u64,
    pub guest_mode: bool,
    pub bot_to_bot: bool,
    pub allowed_bots: Vec<String>,
    pub max_bot_chain_depth: u32,
}

Telegram channel configuration, nested under [telegram] in TOML.

When present, Zeph connects to Telegram as a bot using the provided token. The token must be resolved from the vault at runtime via ZEPH_TELEGRAM_TOKEN.

Example (TOML)

[telegram]
allowed_users = ["myusername"]
stream_interval_ms = 3000
guest_mode = true
bot_to_bot = true
allowed_bots = ["@my_bot"]
max_bot_chain_depth = 1

Fields

token: Option<String>

Bot token. Set to None and resolve from vault via ZEPH_TELEGRAM_TOKEN.

allowed_users: Vec<String>

Telegram usernames allowed to interact with the bot (empty = allow all).

skills: ChannelSkillsConfig

Skill allowlist for this channel.

allowed_tools: Option<Vec<String>>

Tool allowlist for this channel. None means all tools are permitted. Some(vec![]) denies all tools. Some(vec!["shell"]) allows only listed tools.

stream_interval_ms: u64

Minimum interval in milliseconds between streaming message edits.

Defaults to 3000 ms (3 seconds) to stay within Telegram’s rate limits. Values below 500 ms are clamped to 500 ms with a warning; the Telegram Bot API enforces a hard limit of ~30 edits/second per chat.

guest_mode: bool

Enable responding to @mentions in any chat (Bot API 10.0 Guest Mode).

When false (default), guest_message updates are ignored.

bot_to_bot: bool

Enable receiving messages from other bots (Bot API 10.0).

When false (default), messages where from.is_bot = true are silently dropped.

allowed_bots: Vec<String>

Bot usernames allowed to interact when bot_to_bot = true.

Empty list (default) allows all bots. Include the @ prefix (e.g. "@my_bot").

max_bot_chain_depth: u32

Maximum reply chain depth before Zeph stops responding to bot messages.

Prevents infinite loops between bots. Checked against both the structural reply_to_message depth (spec FR-007) and the consecutive-reply counter for the same chat. Default: 1.

Note: Telegram API payloads only expose one level of reply_to_message nesting, so values greater than 1 have no additional effect on structural depth alone. The consecutive-reply counter provides secondary loop prevention across multiple top-level exchanges.

Trait Implementations

impl Clone for TelegramConfig

fn clone(&self) -> TelegramConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for TelegramConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for TelegramConfig

fn deserialize<__D>( __deserializer: __D, ) -> Result<TelegramConfig, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for TelegramConfig

fn serialize<__S>( &self, __serializer: __S, ) -> Result<<__S as Serializer>::Ok, <__S as Serializer>::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.