Skip to main content

Minimal nestable wiki for AI coding harnesses. A folder with `index.md` and a `## Spaces` navigation contract, for any use case (notes, research, recipes, writing, team docs). Spec, conventions catalog, and reference skills. Out of scope: browser-only AI assistants.

Project description

wiki-spaces

A minimal nestable wiki — a folder with index.md and a ## Spaces navigation contract, for any use case. Research, recipes, code notes, writing, team docs, a personal life wiki — your shape, your call.

Markdown flavor is Obsidian — wikilinks, frontmatter, callouts, embeds, comments, Bases. One dialect across the spec, the skills, and the tools. Non-Obsidian renderers (GitHub preview, vanilla VS Code, plain markdown viewers) display the content but will not render Obsidian-specific syntax (provenance comments, embeds, .base files) the way Obsidian does — view your wiki in Obsidian for full fidelity.

Audience

wiki-spaces is built for AI coding harnesses with filesystem access — Claude Code, Codex, Cursor, Windsurf, Gemini CLI, Aider, and similar tools that read and write local files. Browser-only AI assistants (ChatGPT in a tab, Claude.ai web) are out of scope: they can't reach the filesystem to read or write your wiki. If your AI lives in a browser, this isn't the tool.

The shape of the wiki — research notes, recipes, journal, team reference, anything — is yours. The harness that drives it is what wiki-spaces assumes you have.

Install

Installing gives your agent the reference skills plus a scaffolded, registered wiki. Letting an agent do it is the recommended path — setup is a short interview, and an agent runs the steps end to end without fat-fingering a path or a flag.

Let your AI agent do it (recommended)

Paste this to your coding agent (Claude Code, Codex, Cursor, Windsurf, Gemini CLI, Antigravity, …):

Install and set up wiki-spaces for me by following the instructions here:
https://raw.githubusercontent.com/anfreire/wiki-spaces/main/references/SETUP.md

The agent reads SETUP.md, asks what the wiki is for and where it should live, infers a layout, links the skills into your harness, scaffolds the wiki, and writes the config — confirming the plan with you before it runs anything.

Manual

uvx wiki-spaces install                  # link skills into detected harnesses
uvx wiki-spaces init ~/Documents/Wiki              # scaffold a fresh wiki + register it
uvx wiki-spaces init ~/notes --adopt     # OR adopt an existing folder of notes
uvx wiki-spaces doctor --no-net          # verify

install auto-detects and links the skills into whichever of Claude Code, Codex, Gemini CLI, Antigravity, Hermes, and Kiro are present (--all pre-positions for every one). Cursor, Windsurf, GitHub Copilot, and Aider have no global skills directory — they integrate via a rule-file snippet (wiki-spaces install --bridge <key>; see HARNESS_INTEGRATION.md). For a permanent install, pip install wiki-spaces or uv tool install wiki-spaces, then drop the uvx prefix.

init always emits ## Spaces in the new wiki's index.md from t=0 so the space subcommands work immediately. --adopt walks the existing folder, registers every nested folder containing index.md in its ancestor's ## Spaces, and inserts ## Spaces into any pre-existing index that lacks the heading (the v1 navigation contract — a wiki is a folder with index.md AND ## Spaces). After adoption every folder carries both, so audit reports zero drift on day 1. External subtrees (shared/, foreign submodules, escaping symlinks) are skipped with a per-skip notice; pass --include-external to override.

Once a wiki exists, the space subcommands manage its structure:

uvx wiki-spaces space add projects/foo   # create a space + register it
uvx wiki-spaces space audit              # audit drift, broken links, orphans
uvx wiki-spaces space mount <url> --mode submodule              # mount an external space (default path: shared/<basename>/)

space add, space remove, space mount, and space promote auto-insert ## Spaces when missing — the CLI maintains the contract for you. wiki-spaces init writes ## Spaces from t=0 so the first write command lands clean.

No tooling at all

mkdir -p ~/Documents/Wiki && printf '# My Wiki\n\n## Spaces\n\n' > ~/Documents/Wiki/index.md

A folder with index.md + ## Spaces is already a complete wiki — the whole spec is one page, AGENTS.md. The skills still work on it (they discover the wiki from your current directory); run wiki-spaces init later to register it for config-based discovery.

What you get

Three reference skills your AI agent uses to work with the wiki:

  • wiki-search — find content
  • wiki-update — capture / save / sync, with per-file size discipline (hard caps at write time, see CONVENTIONS.md § _meta/limits.md)
  • wiki-tend — audit, normalize tags, cross-link

Cursor / Windsurf / Copilot / Aider integration is covered under Install above and in HARNESS_INTEGRATION.md.

Search at scale

wiki-search works out of the box with grep / ripgrep, which is fine for personal/team wikis up to a few hundred pages. For larger vaults, install qmd — the markdown-aware MCP backend Andrej Karpathy references in the canonical LLM-wiki gist. See CONVENTIONS.md § Recommended search backends.

Learn more

Dependencies

Python >=3.11. uv recommended (handles Python provisioning). git optional.

Prior art

License

MIT. See LICENSE.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

wiki_spaces-1.0.0.tar.gz (131.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

wiki_spaces-1.0.0-py3-none-any.whl (153.6 kB view details)

Uploaded Python 3

File details

Details for the file wiki_spaces-1.0.0.tar.gz.

File metadata

  • Download URL: wiki_spaces-1.0.0.tar.gz
  • Upload date:
  • Size: 131.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wiki_spaces-1.0.0.tar.gz
Algorithm Hash digest
SHA256 e71f132089a9acf4eb93492910e2cf92a6d3fe5343c92a85b116d995f207a72d
MD5 d5f9d839268c0b5a684a4e97e4aefea5
BLAKE2b-256 6fea271c00ff17729894efacc7ca7ccfdcf080bc973b3c38be771bfacd51f6e9

See more details on using hashes here.

Provenance

The following attestation bundles were made for wiki_spaces-1.0.0.tar.gz:

Publisher: release.yml on anfreire/wiki-spaces

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file wiki_spaces-1.0.0-py3-none-any.whl.

File metadata

  • Download URL: wiki_spaces-1.0.0-py3-none-any.whl
  • Upload date:
  • Size: 153.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wiki_spaces-1.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 95964b42861cab525addd0f93c253b8dfbfea637a7a1c35706c3dcc8d1ee7460
MD5 4af72b3442045f064820f2e01f3d3ef5
BLAKE2b-256 d812b98784c3b7bc1bd227f99ad59cdab79ce5d5b17ca2211fdce374c387949d

See more details on using hashes here.

Provenance

The following attestation bundles were made for wiki_spaces-1.0.0-py3-none-any.whl:

Publisher: release.yml on anfreire/wiki-spaces

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page