doc-intelligence-agent

What problem does it solve?

The agent reduces false-positive and context-less vulnerability findings by extracting authoritative product and deployment context from a repository's documentation so security teams can accurately validate, downgrade, or confirm scanner results.

Core Features & Use Cases

• Product context extraction: Summarizes product description, deployment models, and target audience from README and top-level docs.
• Architecture & security mapping: Identifies tech stack, runtime, authentication, authorization, encryption, sandboxing, and operational controls from docs, Dockerfiles, Helm, and Terraform.
• Vulnerability context adjustments: Produces structured doc-profile.json and doc-summary.md to downgrade, confirm, or request more info on scanner findings; useful for generating PR-ready remediation and executive reports.
• Automated + human-in-the-loop: Phases 1–6 automated via scripts/analyze-docs.py; phases 7–8 require agent judgment for external docs and final evidence curation.

Quick Start

Use the doc-intelligence-agent to generate a doc-profile.json and doc-summary.md for the target repository path.