Composio Activity Summary

Company Activity Summary (MCP)

Generate a comprehensive summary of company activity over the last 24 hours (or specified time period) across all connected apps.

Arguments

• $ARGUMENTS - Time period to analyze (default: "last 24 hours"). Examples: "last 24 hours", "last 7 days", "since Monday"

Step 1: Discover Connected Apps

First, check which apps the user has connected by trying lightweight read calls. Use AskUserQuestion if you're unsure which apps they use.

Common apps to check (try each, skip silently if not connected):

Note which apps are connected and skip the rest.

Step 2: Launch Parallel Agents Per Connected App

For each connected app, launch a background Agent to gather and summarize activity. Each agent writes its summary to a markdown file in the current working directory.

Slack Agent

Prompt: Fetch Slack activity for the time period "$ARGUMENTS" (default: last 24 hours).

1. List channels with SLACK_LIST_CHANNELS
2. For active channels, fetch recent messages with SLACK_GET_CHANNEL_MESSAGES_AND_THREAD_REPLIES
3. Focus on: key discussions, decisions made, announcements, questions asked
4. Write summary to slack-activity-summary.md

GitHub Agent

Prompt: Fetch GitHub activity for the time period "$ARGUMENTS" (default: last 24 hours).

1. List repos with GITHUB_LIST_REPOSITORIES_FOR_THE_AUTHENTICATED_USER
2. For each active repo, fetch PRs with GITHUB_LIST_PULL_REQUESTS_IN_A_REPOSITORY
3. For significant PRs, get details and changed files
4. Focus on: PRs opened/merged/reviewed, issues created/closed, code changes
5. Write summary to github-activity-summary.md

Notion Agent

Prompt: Fetch Notion activity for the time period "$ARGUMENTS" (default: last 24 hours).

1. Search recent pages with NOTION_SEARCH_NOTION_PAGE (sort by last_edited_time, descending)
2. For important pages, fetch content with NOTION_FETCH_ALL_BLOCK_CONTENTS
3. Focus on: documentation updates, meeting notes, project tracking changes
4. Write summary to notion-activity-summary.md

Linear Agent

Prompt: Fetch Linear activity for the time period "$ARGUMENTS" (default: last 24 hours).

1. Fetch updated issues with LINEAR_LIST_LINEAR_ISSUES or use LINEAR_RUN_QUERY_OR_MUTATION with GraphQL for richer data
2. Group by team and status changes
3. Focus on: issues completed, issues created, blockers, project progress
4. Write summary to linear-activity-summary.md

Gmail Agent (if connected)

Prompt: Fetch Gmail activity for the time period "$ARGUMENTS" (default: last 24 hours).

1. Fetch recent emails with GMAIL_FETCH_EMAILS
2. Focus on: important threads, action items, external communications
3. Write summary to gmail-activity-summary.md

Calendar Agent (if connected)

Prompt: Fetch Google Calendar activity for the time period "$ARGUMENTS" (default: last 24 hours).

1. List events with GOOGLECALENDAR_LIST_EVENTS
2. Focus on: meetings that happened, key attendees, upcoming schedule
3. Write summary to calendar-activity-summary.md

Step 3: Compile Overview

After all agents complete, read each summary file and present:

# Company Activity Summary - [Date Range]

## Connected Apps Analyzed
- [list of apps that were successfully queried]

## Key Highlights
[3-5 bullet points of the most important things across all apps]

## Detailed Summaries
| Platform | File | Highlights |
|----------|------|------------|
| Slack | [slack-activity-summary.md](./slack-activity-summary.md) | [1-line summary] |
| GitHub | [github-activity-summary.md](./github-activity-summary.md) | [1-line summary] |
| ... | ... | ... |

Read each file and provide a brief overview of key highlights, then let the user know they can open individual files for full details.

Guidelines

• Only summarize apps that are actually connected — don't fail or warn about missing apps
• Use AskUserQuestion if the time period is ambiguous
• For large time periods (>7 days), warn the user it may take longer
• If any agent fails, note which summaries are missing but still present the rest
• Focus on actionable insights, not just lists of items