ProduckAI

Provides multi-source feedback ingestion, AI-powered analysis, VOC scoring, and PRD generation via Claude Desktop.

python

1

GitHub Stars

python

Language

5 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "rohitsaraff33-bit-produckai-mcp-server": {
      "command": "produckai-mcp",
      "args": [],
      "env": {
        "ANTHROPIC_API_KEY": "YOUR_API_KEY_HERE"
      }
    }
  }
}

ProduckAI MCP Server lets you run an AI-powered feedback analysis backend locally and connect it to Claude Desktop. You ingest customer feedback from multiple sources, analyze it with semantic clustering and sentiment signals, score and prioritize insights, and generate executive-ready PRDs all within your workflow.

How to use

You run the MCP server locally and connect it to Claude Desktop. From Claude you issue natural language prompts to upload data, cluster feedback, compute VOC scores, and generate PRDs. Use this flow to triage weekly feedback, prioritize work, and document decisions with evidence.

Typical usage patterns include: ingesting data from Slack, Google Drive, Zoom, JIRA, or CSV uploads; running clustering to identify themes; calculating VOC scores to rank insights; generating PRDs that reference customer quotes; and syncing prioritized work to JIRA.

How to install

Prerequisites: Python 3.11+, Anthropic API key, Claude Desktop.

pip install produckai-mcp-server

Configuration and usage notes

Create a local environment file and point Claude Desktop to the MCP server so it can run commands from Claude prompts.

Create an environment file from a template and add your API key.

Configure Claude Desktop to invoke the MCP server when you start a session.

{
  "mcpServers": {
    "produckai": {
      "command": "produckai-mcp",
      "env": {
        "ANTHROPIC_API_KEY": "YOUR_API_KEY_HERE"
      }
    }
  }
}

First use

In Claude, try these prompts to interact with the MCP server and demo data.

"Upload the demo feedback CSV at ./demo-data/feedback.csv"
"Run clustering and show me the top themes"
"Calculate VOC scores and show top 5 insights"
"Generate a PRD for the highest-priority insight"

Available tools

run_clustering

Generate themes and insights from raw feedback by performing semantic clustering.

generate_embeddings

Create vector embeddings for feedback items to enable semantic search and clustering.

get_themes

List all discovered themes from the clustering process.

get_theme_details

Deep-dive information for a specific theme including related feedback.

search_insights

Natural language search over generated insights.

get_insight_details

Full data for a selected insight including related quotes.

search_feedback

Search raw feedback items by keywords or attributes.

get_customer_feedback

View feedback attributed to a specific customer.

calculate_voc_scores

Compute 0-100 VOC scores for feedback and themes across six dimensions.

get_top_feedback_by_voc

Return a priority-ranked list of insights by VOC score.

configure_voc_weights

Adjust weights for VOC scoring dimensions to tune prioritization.

get_voc_trends

Track VOC score changes over time for insights and themes.

generate_prd

Create a PRD document from a selected insight with executive-ready formatting.

list_prds

Browse generated PRDs.

get_prd

View a full PRD with sections and linked quotes.

update_prd_status

Track PRD workflow stages and status changes.

regenerate_prd

Update a PRD after changes in insights or prioritization.

export_prd

Export PRD content to Markdown for sharing.