Traditional SAST scanners treat your repository as pure code. PullGuard inspects the entire developer agent perimeter: agent instructions (.cursorrules, Claude hooks), Model Context Protocol (MCP) backdoors, and shadow AI dependencies — with zero source code leaving your CI runner.
Attackers poison agent instruction files to trick AI coding assistants into fetching malicious dependencies or exfiltrating tokens.
.cursorrules & .cursor/rules/*.mdc.claude/hooks & desktop configsCustom Model Context Protocol servers give LLMs direct OS and API execution. PullGuard statically audits MCP definitions.
Generate audit-ready evidence for emerging AI safety standards without slowing down engineering momentum.
A real finding, produced by scanning this file with the published PullGuard v1.5.20 image.
PullGuard tracks untrusted user input from HTTP endpoints and webhooks directly into LLM prompts and agent tools in Python, JavaScript/TypeScript, Java, C# and Go.
Python & JavaScript prompt injection sinks & chain invocations
Runner loops, tool dispatch, and unverified action execution
Messages API, tool use schema validation, and credential isolation
streamText, prompt templates, and Java CMS predicate sanitizers
Get started on your public or private repositories for free. No account, no credit card, and zero code uploaded to third-party clouds.