Aguara: Local-First Security Scanner for AI Agents and Software Supply Chains

Aguara is a security scanner developed by garagon, focusing on AI agents and software supply chain integrity. The tool operates locally, eliminating the need for SaaS accounts, telemetry, or external LLM calls. It is designed to detect a range of security threats, including prompt injection, tool poisoning, and risks associated with AI agent configurations (MCP risks). Beyond AI-specific threats, Aguara also scans for vulnerabilities in the broader software supply chain, such as unsafe GitHub Actions, secret exfiltration, and compromised packages across multiple ecosystems including npm, pnpm, PyPI, Go, Rust, PHP, Ruby, Java, and .NET. The scanner reads lockfiles (e.g., `package-lock.json`, `pnpm-lock.yaml`, `yarn.lock`) directly, allowing for checks before package installation. The latest release, v0.22.2, specifically enhances coverage for the Red Hat / Miasma npm compromise, ensuring comprehensive offline detection of affected package versions. Aguara's releases are signed with Cosign keyless via Sigstore, and Docker images include SBOM and SLSA provenance attestations, emphasizing a strong focus on supply chain security for the tool itself.