framework/dist/discord/rest.js

The (AI) Framework: turnkey, zero-config AI orchestration that wraps a coding-agent CLI (Claude Code) as a black box and takes you from an idea to a running app. Vite for AI.

/**
 * The Discord REST calls the bot needs (#680). Separate from the outbound webhook posts in the
 * intervention/activity watchers (#627): a webhook can only speak into one channel and cannot
 * reply, so a bot that answers where it was asked has to go through the API with its token.
 *
 * `fetch` is a parameter, exactly as in the watchers, so tests never touch the network.
 */
const API = 'https://discord.com/api/v10';
/** Discord rejects a message over 2000 characters outright, so a long reply is trimmed. */
export const MAX_CONTENT = 2000;
/** Trim to Discord's limit, marking the cut so a truncated answer never reads as a complete one. */
export function clampContent(text) {
    if (text.length <= MAX_CONTENT)
        return text;
    const notice = '\n… (truncated)';
    return text.slice(0, MAX_CONTENT - notice.length) + notice;
}
/**
 * Post a message to a channel as the bot. Resolves `false` on any failure — a reply that cannot
 * be delivered must never take the daemon down.
 */
export async function postMessage(token, channelId, content, opts = {}) {
    const fetchImpl = opts.fetchImpl ?? fetch;
    const body = { content: clampContent(content) };
    // Thread the answer onto the message that asked, so a busy channel stays readable.
    if (opts.replyToId) {
        body['message_reference'] = { message_id: opts.replyToId, fail_if_not_exists: false };
    }
    try {
        const res = await fetchImpl(`${API}/channels/${encodeURIComponent(channelId)}/messages`, {
            method: 'POST',
            headers: { 'Content-Type': 'application/json', Authorization: `Bot ${token}` },
            body: JSON.stringify(body),
        });
        return res.ok;
    }
    catch {
        return false;
    }
}
//# sourceMappingURL=rest.js.map