speckit-knowledge-pack-generator

To generate a knowledge pack for repository onboarding, the Skill uses script-driven fact collection from the repository root to establish a source-read plan, then applies AI synthesis to create layered markdown guides with source traceability.

AI-assisted knowledge synthesis transforms raw codebase facts into layered markdown guides. It ensures every claim is source-backed, providing traceability from the synthesized knowledge directly back to the original repository files.

Source tracing during project knowledge generation works by linking AI-synthesized claims back to their original repository files. Evidence-backed claim tracing is enforced throughout the quality-loop review to validate coverage and resolve unresolved statements.

Yes, you can validate the shape and coverage of a generated knowledge pack. The process applies quality-loop reviews and equivalence checks to detect missing coverage, unresolved claims, and pack-shape issues before mounting.

No, automated repository analysis replaces manual fact collection with script-driven extraction from the repository root. This establishes a reliable source-read plan and feeds directly into the AI synthesis pipeline without manual hunting.

You should not use automated knowledge pack generation when a repository lacks a clear root directory for script-driven collection, or when you need to manually curate facts without enforced source tracing and equivalence checking.