export.mdx

August 12, 2026 · View on GitHub

Your compiled wiki lives on disk as markdown files, but the knowledge inside it is useful beyond the local viewer and CLI. llmwiki export transforms your wiki into portable formats - a compact llms.txt file that fits in an LLM context window, a structured JSON envelope for programmatic consumers, a JSON-LD graph for semantic tooling, a GraphML file for graph visualization, a Marp slide deck for presentations, or an Open Knowledge Format bundle for exchange with other knowledge tools.

Export is a pure transformation of existing wiki content. It makes no LLM calls and doesn't modify any files in wiki/ or sources/ - only writing artifacts under dist/exports/ or the --out directory you choose for directory-style targets.

Syntax

llmwiki export                                # export default single-file formats
llmwiki export --target <name>                # export one format
llmwiki export --target json --project-id <id>  # export JSON with a stable bridge ID
llmwiki export --target okf --out ./okf-bundle  # export an Open Knowledge Format bundle

When you omit --target, the single-file formats are exported in one pass. The okf target is opt-in because it writes a directory bundle.

Export targets

Target	Flag value	Output file	What it produces
llms.txt	`llms-txt`	`dist/exports/llms.txt`	Concise index per the llmstxt.org spec - page titles, summaries, and links. Compact enough to paste into an LLM context window.
llms-full.txt	`llms-full-txt`	`dist/exports/llms-full.txt`	Full content version of llms.txt - every page's complete body included. Larger but self-contained.
JSON	`json`	`dist/exports/wiki.json`	Structured JSON envelope with per-page metadata (kind, confidence, provenance, citations, aliases, freshness). See JSON export structure below.
JSON-LD	`json-ld`	`dist/exports/wiki.jsonld`	Schema.org JSON-LD graph for semantic web and knowledge graph tooling.
GraphML	`graphml`	`dist/exports/wiki.graphml`	Directed wikilink graph as XML. Import into Gephi, Cytoscape, or any GraphML-compatible tool.
Marp	`marp`	`dist/exports/wiki.md`	Marp presentation slide deck - one slide per concept page.
Open Knowledge Format	`okf`	`dist/exports/okf/`	Directory bundle with `index.md`, per-page OKF docs, copied references, and `log.md`. Use this when another tool expects OKF-style markdown bundles.

Examples

# Export just the llms.txt format
llmwiki export --target llms-txt

# Export the full-content version
llmwiki export --target llms-full-txt

# Export structured JSON
llmwiki export --target json

# Export JSON-LD for a knowledge graph pipeline
llmwiki export --target json-ld

# Export a GraphML graph for Gephi
llmwiki export --target graphml

# Export Marp slides (concept pages only)
llmwiki export --target marp --source concepts

# Export an Open Knowledge Format bundle
llmwiki export --target okf

# Export OKF to a specific directory
llmwiki export --target okf --out ./dist/okf-for-partner

# Export all default single-file formats at once
llmwiki export

`--project-id <id>`

The --project-id flag embeds a stable identifier in the JSON export envelope. Downstream importers use this ID to derive deterministic external IDs for each page - so if you re-export and re-import, pages map to the same records rather than creating duplicates.

llmwiki export --target json --project-id my-research-wiki

Valid project IDs match the pattern /^[a-z0-9][a-z0-9-]{0,62}$/ - lowercase letters, digits, and hyphens, starting with a letter or digit, up to 63 characters. The export is aborted before writing any files if the ID is invalid.

If you're using the [Atomic Memory bridge](/guides/atomic-memory-bridge), always pass the same `--project-id` value on every export. The bridge derives stable memory record IDs from the project ID and page path - changing the ID will create duplicate records on re-import.

JSON export structure

The json target produces a wiki.json file with a top-level envelope and an array of page objects:

{
  "version": 1,
  "projectId": "my-research-wiki",
  "exportedAt": "2026-06-05T09:14:02Z",
  "pages": [
    {
      "path": "concepts/knowledge-compilation.md",
      "slug": "knowledge-compilation",
      "title": "Knowledge Compilation",
      "kind": "concept",
      "summary": "Techniques for converting knowledge representations into efficient forms.",
      "confidence": 0.82,
      "provenanceState": "merged",
      "contradictedBy": [],
      "citations": [
        { "source": "knowledge-compilation.md", "lines": [42, 58] }
      ],
      "aliases": ["knowledge compiler", "compilation"],
      "freshness": "fresh",
      "body": "..."
    }
  ]
}

Per-page fields

Field	Type	Description
`path`	string	Relative path within the wiki (`concepts/` or `queries/`)
`slug`	string	URL-safe page identifier
`title`	string	Page title from frontmatter
`kind`	string	Page kind: `concept`, `entity`, `comparison`, or `overview`
`summary`	string	One-line summary from frontmatter
`confidence`	number \| null	LLM-reported confidence in the synthesized page (0–1)
`provenanceState`	string	`extracted`, `merged`, `inferred`, or `ambiguous`
`contradictedBy`	array	Slugs of pages that contradict this one
`citations`	array	Flattened list of `{ source, lines? }` citations from the page body
`aliases`	array	Alternate names declared in frontmatter
`freshness`	string	`fresh`, `stale`, or `orphaned` based on source state
`createdAt`	string	Creation timestamp from frontmatter; omitted when the page declares none
`updatedAt`	string	Last-updated timestamp, falling back to `createdAt`; omitted when the page declares neither
`body`	string	Full page body (markdown)

The `freshness` field lets downstream consumers filter out stale or orphaned pages before using the content. A page is `stale` when its sources have changed since the last compile; `orphaned` when all its sources have been deleted. Timestamps are read from the page, never from the export run. A page that declares neither `createdAt` nor `updatedAt` — every page written by `llmwiki query --save` declares only the former — exports **no** timestamp rather than an empty one: JSON and JSON-LD omit the keys, GraphML omits the `` elements, and llms.txt, llms-full.txt and Marp omit the `created:`/`updated:` clauses. Two exports of an unchanged wiki are therefore byte-identical.

Marp source filter

When exporting Marp slides, you can narrow the deck to a specific page directory with --source:

llmwiki export --target marp --source concepts  # only concept pages
llmwiki export --target marp --source queries   # only saved query answers
llmwiki export --target marp                    # all pages (default)

When --source narrows the deck, the export summary reports the filtered page count rather than the total wiki size.

Open Knowledge Format export

The okf target writes a directory bundle instead of a single file:

llmwiki export --target okf

By default, the bundle is written to dist/exports/okf/. Use --out <dir> to choose another directory:

llmwiki export --target okf --out ./partner-bundle

The bundle layout is:

dist/exports/okf/
  index.md
  concepts/<slug>.md
  queries/<slug>.md
  <foreign/path>.md
  references/<source-file>.md
  log.md

index.md is the bundle table of contents and carries okf_version: "0.1" in frontmatter.
Native llmwiki pages export under concepts/ and queries/.
Imported foreign pages re-export at their original bundle-relative .md path when that path is safe, URL-safe, non-reserved, and uncontested. Unsafe or colliding paths fall back to the native slug path with a warning.
references/ contains cited source files that still exist under sources/ and resolve inside that directory. Missing or unsafe references are listed as plain text in citations instead of becoming dangling links.
log.md translates llmwiki's activity journal into an OKF-style date-grouped log.

Each page document includes standard OKF fields (type, title, description, tags, timestamp) plus an x-llmwiki block with compiler metadata such as source files, citations, freshness, provenance, and a canonical content hash. Imported foreign OKF pages preserve their foreign type and producer-specific keys on re-export while refreshing the current llmwiki metadata.

type is the page's kind when it declares one, query for pages under wiki/queries/, and concept otherwise. timestamp is the page's updatedAt, falling back to its createdAt; a page that declares neither exports no timestamp at all rather than the time the export ran.

`export --target okf` refuses dangerous output targets such as the filesystem root, the project root, directories inside `.git`, and non-empty directories that are not already OKF bundles. Re-exporting into an existing OKF bundle clears stale bundle files before writing the new bundle.

See Open Knowledge Format guide for the import/export round-trip workflow.

Exporting for the Atomic Memory bridge

The JSON export format is the on-ramp to Atomic Memory. The @atomicmemory/llmwiki bridge imports the wiki.json envelope as one Atomic Memory record per wiki page, preserving all advisory metadata under memory.metadata.llmwiki.*.

# 1. Compile and export
llmwiki compile
llmwiki export --target json --project-id my-research-wiki

# 2. Import into Atomic Memory (using the bridge package)
npx @atomicmemory/llmwiki import dist/exports/wiki.json

See the Atomic Memory bridge guide for the full compile → export → import → package workflow.