source-unifier

Merge documentation from multiple sources by ingesting docs from websites, GitHub repositories, and PDFs into a unified structure. source-unifier combines these sources while detecting conflicts between documentation and code, then outputs a consolidated documentation set with conflict annotations and provenance metadata showing where each piece came from.

Conflict detection identifies mismatches such as missing implementations in code, signature differences between docs and actual code, and outdated documentation. The Skill flags these discrepancies in annotated reports so you can see exactly what conflicts exist and where remediation is needed.

source-unifier supports both rule-based and AI-enhanced merge modes to resolve conflicts. Rule-based strategies apply predefined conflict-resolution logic, while AI-enhanced modes use reasoning to synthesize conflicting information, letting you choose the approach that fits your documentation workflow.

Unified configuration allows you to define merge rules upfront or enable AI-enhanced merging for automatic synthesis. The Skill applies your chosen strategy across all sources, with verification steps built in to ensure the merged output meets your quality standards before generation.

The Skill generates output with unified references and provenance metadata, annotating each section with its source origin—whether from docs-site, GitHub, or PDF. This transparency lets readers trace information back to its original location and assess reliability.

Signature mismatches and missing code implementations are flagged as conflicts during the merge process. The conflict report annotates these discrepancies with remediation guidance, helping you decide whether to update docs, code, or both.