Model Context Protocol · Open Standard

Your org's knowledge,
inside your AI — permanently.

Synaply speaks the open Model Context Protocol — connecting to any MCP-compatible AI assistant and giving it live access to your team's captured intelligence, insight history, and organizational memory.

Overview

What the MCP Integration Does

Most AI assistants start from a blank slate every session. They don't know what your team decided last week, what worked in your last sales call, or what process your ops team documented in March. Synaply's MCP integration fixes this.

Once connected, your AI assistant can read your team's insights, synthesize new ones from your recent work across connected tools, draft structured entries for your review, and query your organizational memory on demand — all from inside your conversation, without switching apps.

The integration exposes 12 tools organized into four areas: Synthesis (generating insights from your recent activity), Read (querying your insight library), Write (creating, editing, and publishing insights), and Identity (understanding your org structure).

Captured once, available everywhere your team works. Synaply doesn't replace your existing tools — it captures the knowledge produced inside them and makes it queryable across your entire AI stack via MCP.

Setup

How to Connect

Synaply works with any MCP-compatible client. The setup below shows the native connector flow — applicable across MCP hosts. You need an active Synaply account. The full setup takes under two minutes.

  1. 1

    Open your AI assistant's settings

    Navigate to Settings → Connectors (or the equivalent MCP server panel) in your AI client.

  2. 2

    Add a New Connector

    Click Add Connector and search for Synaply, or paste the MCP server URL provided by your admin.

  3. 3

    Authenticate via OAuth

    You'll be redirected to Synaply to authorize the connection. Log in with your Synaply credentials and approve the requested permissions.

  4. 4

    Confirm the Connection

    Return to your AI assistant with Synaply listed as active. All 12 tools are immediately available in your conversations.

  5. 5

    Start Your First Session

    Try: "Synthesize my insights from this week." Your assistant pulls your schema, checks your last submission, scans connected sources, and drafts everything for your review.

    Want it to handle this automatically? Download the auto-synthesis skill →
Use Cases

What You Can Do With It

Once connected, these are the most common ways teams use Synaply through their AI assistant.

Weekly Knowledge Synthesis

Your AI scans your connected tools — Slack, email, calendar, Drive — since your last session and auto-drafts structured insights for your review. Takes under a minute.

Synthesize my insights from this week.

Institutional Memory on Demand

Ask your AI anything your team has captured and get answers grounded in real org knowledge — not hallucinated generalities.

What do we know about handling procurement objections?

Team Intelligence Briefing

Pull everything a specific division has captured about a topic, person, or account — filtered by time period, insight type, or author.

Show me Sales insights on outbound messaging from the last 30 days.

Onboarding New Hires

New team members can ask their AI to surface everything the team knows about their role, accounts, or processes — before their first meeting.

What should I know about our enterprise sales process?

Leadership Analytics

Get submission trends, participation rates, and insight type breakdowns across your divisions — useful for quarterly reviews and identifying knowledge gaps.

Summarize Q1 analytics — who's active and what's missing?

Structured Manual Capture

Walk your AI through what you learned in a meeting, call, or project — it formats it into your org's insight schema and saves it as a draft for review.

Capture a process insight from my call just now.
API Reference

All 12 Tools

Every tool available through the Synaply MCP server. Required parameters marked .

get_insight_schemaSynthesis

Returns the full insight schema for all divisions — every insight type, guideline, and form field. Called first at the start of every synthesis session.

get_last_submissionSynthesis

Returns when you last published an insight per division. Sets the lookback window for synthesis so your AI only surfaces what's new since your last session.

get_insightsRead

Fetches insights you have access to. Filterable by division, insight type, author, date range, and location. Use this to browse or pull insights in bulk.

search_insightsRead

Full-text search across your insight library by keyword. Use this to answer "what do we know about X?"

Required:query
get_division_insightsRead

Returns the insight types available in a division along with their form fields and guidelines. Used before creating an insight manually.

Required:divisionId
create_insightWrite

Creates a new insight entry and saves it as a draft. Works automatically during synthesis or interactively in manual mode.

Required:contenttitle
update_insightWrite

Edits an existing insight — content, type, title, or privacy level.

Required:entryIdcontent
publish_insightWrite

Publishes a draft insight to your division. Only called after you explicitly confirm — your AI never publishes automatically.

Required:entryId
get_draftsRead

Returns all unpublished drafts saved on your behalf. Useful for resuming a session or reviewing insights before publishing.

get_user_identityIdentity

Returns your profile and org context. Supports six query types: identity, divisions, team_structure, locations, manager, org_chart.

Required:queryType
search_usersIdentity

Looks up a team member by name and returns their user ID. Required for filtering insights or analytics by a specific person.

Required:name
get_analytics_summaryRead

Returns aggregated analytics across your team — submission trends, participation rates, and insight type breakdowns. Filterable by date range and insight type.

Reference

Privacy Levels

Every insight has a privacy level controlling who can read it. Your AI defaults to EVERYONE during auto-synthesis. Specify a different level explicitly at any point if needed.

ONLY_ME

Visible only to you — private notes and personal captures.

DIRECT_SUPERVISOR

You and your direct manager.

ALL_SUPERVISORS

You and your full upward reporting chain.

MY_TEAM

You and everyone who shares your manager (your peer group).

MY_DIVISION

All members of your division.

EVERYONE

All authenticated users in your organization. Default for auto-synthesis.

Security

How the Integration Is Secured

Synaply's MCP integration is built with security as a baseline, not an afterthought.

OAuth 2.0 authentication

All connections are established via standard OAuth 2.0. No passwords are exchanged — your AI assistant receives a scoped access token that can be revoked at any time from your Synaply account settings or from the assistant's connector panel.

Encrypted transit

All data exchanged between Synaply's servers and your AI assistant provider travels over encrypted HTTPS connections.

Scoped access

The MCP integration only accesses data your account is already authorized to see — existing role-based permissions and insight privacy levels are enforced on every tool call. An insight marked ONLY_ME is never returned to another user's session.

No AI training

Data accessed through the MCP integration is not used to train AI models, by Synaply or any third party.

Revocation

Disconnect the Synaply connector from your AI assistant's settings at any time. This immediately terminates the integration's access with no further action required.

Support

Troubleshooting

My AI assistant says Synaply tools aren't available+

The connector may have been disconnected or the OAuth session expired. Go to your assistant's connectors panel and confirm Synaply shows as active. If listed but unresponsive, disconnect and reconnect — the OAuth flow issues a fresh token.

Auto-synthesis returns no drafts+

Either (a) no signal-producing connectors are enabled — try connecting Slack, Gmail, or Google Calendar — or (b) there's nothing new since your last submission. Synthesis only covers activity after your last submission timestamp per division.

Insight types are missing or wrong+

Insight types are configured by your workspace admin in Synaply. If types are missing for your division, contact your admin. Your AI reads the live schema every session — nothing to refresh on the client side.

How do I revoke the connection?+

Disconnect the Synaply connector from your AI assistant's connectors panel. This immediately revokes the token and terminates access. You can also contact support@synaply.io to force-revoke all active tokens for your account.

Still stuck? Email support@synaply.io with your workspace name and a description of what you're seeing.

Go Further

Auto-Synthesis Skill

The MCP integration gives your AI the tools. The auto-synthesis skill tells it exactly how to use them — scanning your connected work, identifying what's worth capturing, drafting structured insights, and walking you through approval. All from a single prompt.

Synaply Auto-Synthesis Skill
Works with any MCP-compatible AI · Requires Synaply MCP connected

Teaches your AI to run your full weekly synthesis session end-to-end — schema lookup, lookback window, connector scanning, draft generation, and publish flow. Just say “synthesize my insights” and it handles the rest.

Download Skill

To install: download the .skill file and upload it to your AI assistant's skills folder, or follow your team's skill installation process. The MCP connection must be active for the skill to work.