Access
Read only
Feature
VueLeaf MCP is a read-only connection to your workspace. Ask what changed, why, and where, and get an answer that shows its scope and the posts behind it.
Access
Read only
Workspace per answer
1
Data tools
9
Ready-made prompts
33
Why did our sentiment drop this week?
get_sentiment_reportget_sentiment_attribution
Negative share rose on two grower forums. Most of the shift traces to one troubleshooting thread about a product defect.
One source had low volume in this window, so its share is directional.
Supporting posts returned with forum, author, and date so the team can open the threads.
What it answers
Each question maps to the VueLeaf tools the assistant calls to answer it.
Use the daily brief and smart insights to summarize the newest movement before a leadership update, retailer call, or internal stand-up.
get_daily_briefget_smart_insights
Pull share of voice and market momentum into one conversation so the team can compare its position with tracked competitors without rebuilding the analysis manually.
get_share_of_voiceget_market_momentum
Use the sentiment report, sentiment attribution, and source breakdown to see which forums, topics, or author groups pushed the change.
get_sentiment_reportget_sentiment_attributionget_source_breakdown
Search mentions directly from the assistant to pull the supporting threads when someone asks for examples, proof, or edge cases.
search_mentions
Pull the current reputation risk score, the contributing signals, and the underlying evidence so the team can decide what to inspect next.
get_reputation_risk
Setup
There is no shared key to manage. Each person signs in with their own VueLeaf account, so access follows the permissions they already have.
01
In an AI assistant that supports MCP connections, add https://vueleaf.com/mcp as a connector.
02
The assistant gets read-only access to your current workspace and only sees what you can already see in VueLeaf.
03
Start with what changed in a set window, check the scope on the answer, and ask for the supporting posts when the answer needs proof.
Provenance
The assistant does not just return a conclusion. Each VueLeaf tool result carries the context your team needs to check it, so nobody has to take a summary on faith.
That is the difference between pasting an export into a chat and asking VueLeaf directly.
How VueLeaf scores and attributes data| Field | What it tells you |
|---|---|
| Workspace | Which workspace and brand set the answer came from |
| Date window | The exact period, so a 7-day read is not mistaken for a quarter |
| Freshness | When the underlying data was last refreshed |
| Denominator | How many posts or mentions the figure is based on |
| Warnings | Low volume, partial coverage, or other caveats |
| Evidence | The posts behind the conclusion, when requested |
Ready-made prompts
Teams that do not want to write prompts can run a saved one. Most come with a preset 7, 30, or 90 day window or an advanced version for exact dates and filters, for 33 prompts in total.
Boundaries
Read only
The assistant cannot change keywords, alerts, or reports through MCP. It reads and explains.
One workspace
Each answer is scoped to the workspace you are signed in to. Data from other brands or teams never mixes in.
Human review
Answers speed up analysis and prep. Your team still reads the evidence before acting on it.
Not a report
Use live questions for investigation. Use VueLeaf reports when you need a fixed artifact to share.
Go deeper