Code scanning
Scan supplied code with the rules supported by the connected server.
Use a connected Semgrep MCP server to inspect code in qbash. Ask the model to explain the scan findings and point to the affected code.
MCP server
Give ChatGPT, Claude and your other AI tools access to SemGrep through one server. You pick which SemGrep tools it shares, and you can add other connectors, memory folders and Tasks to the same server.
qbash
Supply the code you want checked and ask for findings tied to specific lines.
Scan this Python handler with Semgrep. For each finding, show the rule, affected line, and a proposed fix. Keep scan findings separate from additional suggestions.
Scan supplied code with the rules supported by the connected server.
Explain reported issues and inspect suggested changes with their rule context.
Scan a request handler and discuss the reported findings before editing it.
Use a server with custom-rule support to check whether a rule matches the supplied code sample.
Authentication depends on the Semgrep server you connect. Access to Semgrep AppSec Platform findings requires a Semgrep token.
Open the SemGrep connector in qbash and check its connection settings. For a custom remote server, supply its HTTP or SSE URL and the authentication required by its operator.
MCP connectorsSemgrep does. The server now ships in the main semgrep binary, and Semgrep also runs a hosted beta at mcp.semgrep.ai/mcp that it describes as experimental. qbash connects to whichever of these you configure.
No. Semgrep has no methods in the qbash Task SDK, so its tools run from qbash.
Connect the server and enable its tools in qbash.