searchLABSCO SUMMARY
Two tools, and a deliberate two-step: titles first, then the report you actually want.
Search returns titles rather than full text, which keeps a broad query from flooding the context — you pick one and fetch it. That makes it usable during a review, where the question is usually whether this class of bug has been written up before.
What it is
A small server over Solodit's vulnerability reports: search by keyword, then fetch a report whole.
What you get
- search takes keywords and returns the matching report titles and slugs
- get-by-slug returns the full content of one report, by the slug search gives you
- Runs over HTTP at POST /mcp, so it can serve more than one client
Requirements
Nothing to sign up for. Start it with npx and the @lyuboslavlyubenov/solodit-mcp package, or run the published Docker image on port 3000; clients point at http://localhost:3000/mcp. A local build with pnpm is the third path.
Setup effort
One command — npx @lyuboslavlyubenov/solodit-mcp
USE CASES
01
Checking whether a bug pattern has been found before
Search by keyword and scan the report titles that come back.
02
Reading the full write-up of a finding
Fetch the report by the slug search returned and go through it in the thread.
03
Gathering prior art for an audit report
Search the area, then pull the reports worth citing one at a time.
2 tools
Listed in project code.
get-by-slugBEFORE YOU INSTALL
What to know first
Tools you get
2 tools registered in the project's code. The README calls the second tool get-by-title; the server registers it as get-by-slug, and it takes the slug that search returns.
Requires
npx on PATH
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
TERMINAL
npx @lyuboslavlyubenov/solodit-mcpMCP Clients
Any MCP client
ANY MCP CLIENT · JSON
{
"mcpServers": {
"solodit-mcp": {
"url": "http://localhost:3000/mcp"
}
}
}
