ai-shipping-labs-slack

name	ai-shipping-labs-slack
description	Use when asked to read AI Shipping Labs Slack — get a Slack user's email, resolve a Slack handle/display name/real name to an email, look up a Slack user by ID, read a Slack thread or channel, or answer "who posted in
metadata	{"short-description":"Read Slack users and public messages; resolve handles to emails"}

Slack: read users and messages

Read-only access to the AI Shipping Labs Slack workspace: resolve people to emails and read public threads/channels. This skill is Slack-only. Writing CRM records is the ai-shipping-labs-users skill; the end-to-end thread->CRM recipe is slack-thread-to-crm.

Credentials

SLACK_BOT_TOKEN lives in the repo .env. Read it inline at call time; NEVER print, paste, or commit it.

cd /home/alexey/git/ai-shipping-labs
SLACK_TOKEN=$(grep -E '^SLACK_BOT_TOKEN=' .env | head -1 | cut -d= -f2- | tr -d '"' | tr -d "'" | tr -d '\r')

Header: Authorization: Bearer <token>.

Scopes available: users:read, users:read.email, channels:read, channels:history, chat:write. Hard limits that follow:

• Public channels and public history only.
• User profiles and emails are readable.
• NO private channels: groups:read / groups:history are not granted. Passing private_channel to conversations.list returns missing_scope.

Resolve a person to an email

By Slack user ID (users.info)

curl -s -H "Authorization: Bearer $SLACK_TOKEN" "https://slack.com/api/users.info?user=$U" \
  | python3 -c "import sys,json;u=json.load(sys.stdin)['user']['profile'];print(u.get('real_name'),u.get('email'))"

By display name / real name (users.list)

users.info needs an ID. When you only have a handle or a name from a screenshot, page the directory and match yourself. users.list returns up to limit members per page; follow response_metadata.next_cursor until it is empty. Match against name (handle), profile.real_name, and profile.display_name.

import json, subprocess

TOKEN = subprocess.run(
    "grep -E '^SLACK_BOT_TOKEN=' .env | head -1 | cut -d= -f2- | tr -d '\"'\"'\"'\\r'",
    shell=True, cwd="/home/alexey/git/ai-shipping-labs",
    capture_output=True, text=True).stdout.strip()

needle = "ostrovnoy"
cursor = ""
while True:
    url = f"https://slack.com/api/users.list?limit=200&cursor={cursor}"
    out = subprocess.run(
        ["curl", "-s", "-H", f"Authorization: Bearer {TOKEN}", url],
        capture_output=True, text=True).stdout
    data = json.loads(out)
    for m in data.get("members", []):
        p = m.get("profile", {})
        hay = " ".join(filter(None, [
            m.get("name"), p.get("real_name"), p.get("display_name")])).lower()
        if needle in hay:
            print(m["id"], m.get("name"), p.get("real_name"), p.get("email"))
    cursor = data.get("response_metadata", {}).get("next_cursor", "")
    if not cursor:
        break

This is exactly how Ostrovnoy resolved to [e-mail] (real_name "Kir Ostrovnoy", id U0ATQL8MUUE).

Reverse direction (users.lookupByEmail)

When you already have an email and want the Slack user, use users.lookupByEmail?email=... instead of paging the whole directory.

Find a channel

curl -s -H "Authorization: Bearer $SLACK_TOKEN" \
  "https://slack.com/api/conversations.list?types=public_channel&limit=1000" \
  | python3 -c "import sys,json;[print(c['id'],c['name']) for c in json.load(sys.stdin)['channels'] if 'community' in c['name'].lower()]"

The main channel is #community = C0AFZSRAYQ4. The bot must be a member of the channel (is_member=True) to read its history.

Read a thread

Find the parent message in recent history, then pull its replies. Note the parent ts and reply_count.

curl -s -H "Authorization: Bearer $SLACK_TOKEN" \
  "https://slack.com/api/conversations.history?channel=$CH&limit=200"

curl -s -H "Authorization: Bearer $SLACK_TOKEN" \
  "https://slack.com/api/conversations.replies?channel=$CH&ts=$TS&limit=200"

reply_count includes the parent message. Read every reply yourself; do not regex the text for intent.

Next steps

• Resolving Slack people to platform accounts and writing CRM records: ai-shipping-labs-users.
• The full thread->CRM recipe (read thread, classify intent, write notes + tags): slack-thread-to-crm.