Prior art

Project history

Origins: Obsidian Frameworker

The project began as “Frameworker” — a concept for mapping plaintext markdown files to taxonomical frameworks with CSV import, YAML frontmatter, and 2-way sync. The core insight: GRC involves massive knowledge work, and keeping structured frameworks side-by-side with the knowledge we create could change how compliance teams operate.

Python tool (1500+ lines)

The frameworks_to_obsidian.py script was the first implementation. It takes tabular framework data (CSV/XLSX) and translates it into hierarchical Obsidian vault structures with:

• Folder taxonomy generation from hierarchy columns
• Markdown notes with YAML frontmatter
• WikiLinks between framework nodes (crosswalks)
• Custom transforms per framework (tag aggregation, regex normalization, hierarchical forward-fill)

Supported frameworks: MITRE ATT&CK + D3FEND + ENGAGE, NIST 800-53, CSF v2, CIS v8, CRI Profile

Architecture: Load → Clean → Transform → Deduplicate → Configure → Generate → Link

Obsidian plugin

The plugin extends the Python tool into a general-purpose import wizard with a UI, config system, and broader format support. It makes the import process accessible to non-developers.

Obsidian & Cyber Working Group

An open-source community building resources for using Obsidian in cybersecurity:

• Framework crosswalk engine
• Starter vaults for SecOps, GRC, and personal branding
• Knowledge organization meta-framework SEACOW — lives in the cybersader/cyberbase public vault. See the SEACOW + folder-tag-sync prior-art log for how it connects to Crosswalker’s Foundation decisions.
• obsidian-folder-tag-sync — the project owner’s other active plugin (lives in the same plugin_development/ workspace as Crosswalker) that provides bidirectional folder↔tag sync for polyhierarchy. Direct reuse candidate for Crosswalker’s generation engine — see the prior-art log and the roadmap research item.

The edge metadata gap

A key discovery during research: no existing Obsidian plugin can attach data to relationships between notes. Every plugin evaluated falls short:

Plugin	What it does	Why it’s not enough
Breadcrumbs V4	Edge fields and field groups	Individual key:value pairs per edge only — no structured JSON metadata
Juggl	Graph visualization with typed links	No powerful query/reasoning over the graph; needs export/import
Excalibrain	Visual graph like Juggl/Breadcrumbs	Display-only, no metadata encoding
Graph Link Types	Turns Dataview inline tags into graph edges	Display-only, no additional metadata
Semantic Canvas	Another graph view with link types	Display-only
Supercharged Links	CSS styling based on target note YAML	Visual only, no edge data

This gap is what motivated the link metadata system — using Dataview inline fields with dot notation and JSON to encode structured metadata on links.

Existing tools in this space

GRC platforms

• ServiceNow GRC, Archer, LogicGate — Enterprise GRC with proprietary databases
• Drata, Vanta — Automated compliance, SaaS-locked
• CISO Assistant — Open-source, 80+ frameworks, auto-mapping between standards
• Trade-off: Full-featured but expensive (commercial) or limited UI (open-source), locked-in

Framework mapping tools

• SCF with STRM — Set Theory Relationship Mapping, mathematical approach to crosswalking
• Apptega — Commercial crosswalking platform
• HITRUST MyCSF — Compliance framework tool
• RegScale — OSCAL-native compliance automation
• See framework standards for full list

Knowledge/note tools

• Obsidian + Dataview — Flexible but no structured import
• Notion databases — Good structure but no graph navigation
• Roam Research — Graph-first but no filesystem

Gap

No tool bridges structured framework data and a personal/team knowledge base with typed, queryable links. That’s the Crosswalker niche.

Related Obsidian plugins

Plugin	Overlap	Difference
Dataview	Queries generated data	Doesn’t import structured data
Folder Notes	Folder-as-note pattern	No import or generation
Metadata Menu	Property management	No bulk import from external files
Database Folder	Table views	No external data import wizard
Obsidian Bases	Built-in table/card views	No external data import

Python tool technical details

Helper functions

• match_value() — flexible string/array/regex matcher for crosswalking
• build_full_path_components() / split_folders() — path construction
• sanitize_column_name() — normalize column headers
• hierarchical_ffill() — grouped forward-fill for merged cells in Excel

Core engine

• Taxonomy builder — DataFrame + config → folders + markdown files
• Linker — two frameworks + matching rules → inject WikiLinks

Per-framework transforms

• CRI: Tag aggregation from subject areas
• NIST 800-53: Regex ID normalization for assessment procedures
• D3FEND: Hierarchical forward-fill for ontology structure
• CIS v8: Preamble extraction from control descriptions

See helper functions for algorithm specifications and framework data sources for per-framework sheet and column details.

Cross-project learnings: DFD-Excalidraw System

The DFD-Excalidraw System is a separate project that transforms Excalidraw diagrams into a queryable markdown-based graph database — solving identical problems to Crosswalker:

Shared Problem	DFD Solution	Crosswalker Application
File-based graph database	Markdown files as nodes (Assets) and edges (Transfers)	Framework notes as nodes, WikiLinks as edges
No referential integrity	Defensive scripts that self-heal on each run	Import wizard validates before generating
Schema versioning	dfd-asset-v1, dfd-transfer-v1 identifiers	_crosswalker.schema_version field
Orphan detection	Forward scan + Bases audit views	Backlink-count detection via Bases
Data model resilience	4-tier metadata system	_crosswalker metadata tiers
Eventual consistency	Manual sync via script runs	Manual sync via re-import

Key insight from DFD: design for lazy detection, not eager prevention. File-based systems can never be ACID-compliant, so the focus should be on surfacing inconsistencies (via Bases views) and providing repair tools (re-import, link cleanup), not trying to prevent all invalid states.