Pages API

Collaborative documents that become searchable organisational knowledge the moment they are saved.

Chordian Pages is a document workspace whose pages are written into Chordian Memory. A page you save is chunked, embedded, entity-extracted into the knowledge graph, and synthesised — so what your team writes becomes searchable organisational knowledge by default, not by a separate export step.

Two planes, and why the split matters

The API is split across two surfaces. The split is a data-residency guarantee, not an implementation detail.

PlaneServesRuns on
Metadata — this APIpage tree, projects, onboarding, searchChordian platform
ContentMemory APIMarkdown bodies, versions, assetsthe memory that owns the page

Page bodies never transit the Chordian platform. On a sovereign deployment they are written directly to your own infrastructure, and the platform sees only identifiers, titles, timestamps and sync state. That property is what makes Pages usable by customers who cannot let document content leave their environment.

Where a page lives

Pages inherits the memory model rather than inventing one.

Memory typeMarkdown filesMetadataVectors & graph
Shared secure memoryChordian-managed storage, tenant-isolatedChordian platformshared stores
Dedicated memoryyour boxChordian platformyour box
Sovereign (your infrastructure)your box — never leavesyour boxyour box

Air-gapped deployments run Pages fully offline.

Storage format

A page is Markdown with YAML frontmatter:

---
page_id: pg_7f3a91c2
project_id: prj_a1b2c3d4
title: Q3 Architecture Decisions
version: 7
format: md
---

# Q3 Architecture Decisions

Body content in standard Markdown…

It is readable and editable with no Chordian tooling — no proprietary wrapper, no base64 blob. Exporting a project yields a zip that preserves the page tree as nested folders. This is a product guarantee: your documents remain yours, in a format that outlives any vendor.

Next

Was this page helpful?
Report an issue

On this page