generate_manager_tearsheetLABSCO SUMMARY
A diligence tearsheet, generated rather than assembled.
The pass an analyst makes over a manager — filings, performance, service providers, risk flags — comes back as one structured PDF instead of a folder of tabs.
What it is
A research pipeline exposed as an MCP server. An agent fleet gathers from SEC EDGAR, filings, news and the web through Firecrawl; an assembler normalizes the result into fund profile, performance, risk signals and service providers; ReportLab formats it as a tearsheet.
What you get
- generate_manager_tearsheet produces the full tearsheet for a named manager
- list_available_managers shows who can be covered
- get_manager_coverage reports what is held on a given manager
- get_pipeline_status reports on a run in progress
- Output follows a diligence framework encoded from institutional buy-side practice, with a sample tearsheet in the repository
Requirements
Nothing local: it runs as a hosted MCP server, callable from Claude Desktop, Cursor or any other MCP client. Behind it sit Python, the Anthropic and OpenAI APIs, Qdrant, n8n, Firecrawl and ReportLab.
Setup effort
One command — smithery mcp add ndatta18-b06a/altbots
USE CASES
Preparing for a first meeting with a manager
The profile, performance and service-provider picture arrive in one document, so the meeting starts from questions rather than basics.
Checking whether a manager is covered at all
Coverage is answered before a request is made, which saves waiting on a run that has nothing to work from.
Following a report that is still generating
Pipeline status while the agents work, so a long run is tracked rather than guessed at.
Building a comparable set across several managers
The same framework applied to each name means the tearsheets line up against each other rather than needing normalizing.
4 tools
Listed in project docs. Your client may show a different list.
get_manager_coveragelist_available_managersget_pipeline_statusBEFORE YOU INSTALL
What to know first
Tools you get
4 tools named in the project's documentation.
Requires
A client that can reach a remote MCP server over HTTP
INSTALLATION
Connect your client
Server setup
The command line the project publishes
Smithery CLI
smithery mcp add ndatta18-b06a/altbotsHosted endpoint
https://altbots--ndatta18-b06a.run.toolsAdd this hosted endpoint to your MCP client.