i9wa4

slack

Fetch Slack thread from URL using API directly (no MCP server needed). Use when: - User provides a Slack message URL - User says "read this slack thread" or "fetch slack thread" - User wants to see conversation context from Slack - User wants to search Slack messages - User wants to fetch channel history (e.g., Google Calendar DM)

i9wa4 12 Actualitzat fa 7 mesos
GitHub

Instal·la

npx skillscat add i9wa4/dotfiles/slack

Instal·la des del registre de SkillsCat.

SKILL.md

Slack API Skill

Fetch threads and search messages using direct API calls (no MCP server needed).

1. Prerequisites

1.1. Environment Variables

export SLACK_MCP_XOXC_TOKEN="xoxc-..."
export SLACK_MCP_XOXD_TOKEN="xoxd-..."

2. URL Formats

2.1. Message URL (archives format)

https://xxx.slack.com/archives/C1234567890/p1234567890123456
  • C1234567890 = channel ID
  • p1234567890123456 = timestamp (remove p, insert . before last 6 digits)
    • p1234567890123456 → 1234567890.123456

2.2. Channel URL (client format)

https://app.slack.com/client/E1234567890/C1234567890
https://app.slack.com/client/E1234567890/D1234567890
  • First ID (E...) = Enterprise/Workspace ID
  • Second ID = channel ID (C = channel, D = DM)

3. Fetch Channel History

Fetch latest messages from a channel (useful for bot DMs like Google Calendar).

3.1. Extract Channel ID from Client URL

URL="https://app.slack.com/client/E05CUN3JKJN/D07UW8G7C9H"
CHANNEL=$(echo "$URL" | sed -n 's|.*/client/[^/]*/\([^/]*\).*|\1|p')
echo "$CHANNEL"  # D07UW8G7C9H

3.2. Fetch Latest Messages

CHANNEL="D07UW8G7C9H"
FILE=$(mkoutput --dir tmp --label output)
curl -s -X POST "https://slack.com/api/conversations.history" \
  -H "Authorization: Bearer $SLACK_MCP_XOXC_TOKEN" \
  -H "Cookie: d=$SLACK_MCP_XOXD_TOKEN" \
  -d "channel=${CHANNEL}" \
  -d "limit=20" > "$FILE"

# Extract calendar events (for Google Calendar DM)
jq '.messages[] | select(.attachments) | .attachments[] | select(.title) | {pretext, title, text}' "$FILE"

4. Fetch Thread

4.1. Parse URL and Fetch

# Example URL: https://xxx.slack.com/archives/C1234567890/p1234567890123456
CHANNEL="C1234567890"
TS="1234567890.123456"

curl -s "https://slack.com/api/conversations.replies?channel=${CHANNEL}&ts=${TS}" \
  -H "Authorization: Bearer ${SLACK_MCP_XOXC_TOKEN}" \
  -H "Cookie: d=${SLACK_MCP_XOXD_TOKEN}" | jq '.messages[] | {user, text, ts}'

4.2. One-liner with URL Parsing

URL="https://xxx.slack.com/archives/C08V9MYDQ1C/p1736929199388759"
CHANNEL=$(echo "$URL" | sed -n 's|.*/archives/\([^/]*\)/.*|\1|p')
TS_RAW=$(echo "$URL" | sed -n 's|.*/p\([0-9]*\).*|\1|p')
TS="${TS_RAW:0:10}.${TS_RAW:10}"
curl -s "https://slack.com/api/conversations.replies?channel=${CHANNEL}&ts=${TS}" \
  -H "Authorization: Bearer ${SLACK_MCP_XOXC_TOKEN}" \
  -H "Cookie: d=${SLACK_MCP_XOXD_TOKEN}" | jq '.messages[] | {user, text, ts}'

4.3. Fetch Single Message (conversations.history)

CHANNEL="C1234567890"
TS="1234567890.123456"

curl -s -X POST "https://slack.com/api/conversations.history" \
  -H "Authorization: Bearer ${SLACK_MCP_XOXC_TOKEN}" \
  -H "Cookie: d=${SLACK_MCP_XOXD_TOKEN}" \
  -d "channel=${CHANNEL}" -d "latest=${TS}" -d "limit=1" -d "inclusive=true" | jq '.messages[0]'

5. Output Format

Show each message with:

  • User ID (or resolve to name if needed)
  • Message text
  • Timestamp

6. Search Messages

NOTE: Enterprise Grid may restrict search.messages API
with enterprise_is_restricted error.

6.1. Basic Search

QUERY="keyword"
curl -s "https://slack.com/api/search.messages?query=$(echo "$QUERY" | jq -sRr @uri)&count=20" \
  -H "Authorization: Bearer ${SLACK_MCP_XOXC_TOKEN}" \
  -H "Cookie: d=${SLACK_MCP_XOXD_TOKEN}" | jq '.messages.matches[] | {channel: .channel.name, user, text, ts, permalink}'

6.2. Search with Filters

# Search in specific channel
QUERY="in:#channel-name keyword"

# Search from specific user
QUERY="from:@username keyword"

# Search with date range
QUERY="after:2026-01-01 before:2026-01-15 keyword"

# Combined filters
QUERY="in:#general from:@john after:2026-01-01 terraform"

curl -s "https://slack.com/api/search.messages?query=$(echo "$QUERY" | jq -sRr @uri)&count=20&sort=timestamp&sort_dir=desc" \
  -H "Authorization: Bearer ${SLACK_MCP_XOXC_TOKEN}" \
  -H "Cookie: d=${SLACK_MCP_XOXD_TOKEN}" | jq '.messages.matches[] | {channel: .channel.name, user, text, ts, permalink}'

6.3. Search Options

  • count: Number of results (default 20, max 100)
  • sort: score (relevance) or timestamp
  • sort_dir: asc or desc

7. Error Handling

  • If ok: false in response, show error message
  • Common errors:
    • channel_not_found - invalid channel ID
    • thread_not_found - invalid timestamp
    • invalid_auth - token expired or invalid
    • enterprise_is_restricted - Enterprise Grid API restriction