Skip to main content

Semantic search CLI for markdown vaults

Project description

memex-md

You like Obsidian? Your LLM will love it too.

Memex: Vannevar Bush's 1945 concept of a "memory extender" - a device for storing and retrieving personal knowledge. The conceptual ancestor of personal wikis and second brains.

Semantic search and wikilink graph traversal for markdown vaults. Uses any sentence-transformers model. Point it at your Obsidian vault (or any markdown folder).

Quick Start

uvx memex-md --help

or

uv tool install memex-md

If installed, you can run memex directly, or via the alias mx:

mx vault:add personal ~/notes ~/journal
mx search "How does the auth flow handle token refresh?" -v personal
mx explore auth personal

For a memex skill, see my agent workflows.

How It Works

A vault is a named collection of directories. Each vault has its own embedding model and SQLite index (~/.local/share/memex-md/<vault-name>/index.db).

The index contains:

  • Embeddings for semantic similarity (default: embeddinggemma-300m)
  • Wikilink graph for backlink/outlink queries
  • Extracted frontmatter (aliases, tags)

Indexing is incremental — on each command, only files with changed mtimes are re-indexed. Hidden directories (.obsidian, .trash, .git, etc.) are excluded.

Note: Initial indexing requires embedding computation. Example: ~3800 notes took ~7 minutes on an RTX 3070 Ti.

Configuration

Vaults are configured via CLI. Config lives at ~/.config/memex/config.toml.

mx vault:add personal ~/notes ~/journal
mx vault:add work ~/work-docs --model some/other-model
mx vault:list
mx vault:info personal
mx vault:remove personal --path ~/journal   # remove one path
mx vault:remove work                        # remove entire vault

Re-add a vault with --model to change its embedding model (triggers re-embedding). Use --model none to disable semantic search (wikilink navigation only).

Commands

search

mx search "How does the auth flow handle token refresh?" -v personal
mx search "What approaches did we consider for caching?" -v work --full

Embeds the query and ranks indexed notes by cosine distance. Natural language questions of a few sentences tend to work well.

Flag Description
-v, --vault Search a specific vault (default: all)
-n, --limit Max results (default: 5)
-p, --page Pagination (default: 1)
-f, --full Include note content (default: paths only)

explore

mx explore auth personal
mx explore docs/api-design work --full

Shows a note's outlinks ([[wikilinks]]), backlinks, and semantically similar notes.

note_path can be a title (auth) or path (docs/auth.md). Titles must be unique in the vault.

Flag Description
-f, --full Include note content and metadata (default: graph only)

rename

mx rename old-name new-name personal
mx rename docs/guide manual work

Renames a note file and updates all [[wikilinks]] pointing to it. Handles path links, title links, aliases, and heading refs. Ambiguous links (multiple files share a name) are skipped with warning.

index

mx index
mx index -v personal

Trigger indexing. Runs automatically before search/explore.

Development

uv sync
make check          # ruff + ty
make test           # pytest
make release-patch  # 1.0.0 -> 1.0.1, tag, push
make release-minor  # 1.0.0 -> 1.1.0
make release-major  # 1.0.0 -> 2.0.0

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

memex_md-2.0.4.tar.gz (16.8 kB view details)

Uploaded Source

Built Distribution

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

memex_md-2.0.4-py3-none-any.whl (20.2 kB view details)

Uploaded Python 3

File details

Details for the file memex_md-2.0.4.tar.gz.

File metadata

  • Download URL: memex_md-2.0.4.tar.gz
  • Upload date:
  • Size: 16.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.12 {"installer":{"name":"uv","version":"0.9.12"},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"22.04","id":"jammy","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for memex_md-2.0.4.tar.gz
Algorithm Hash digest
SHA256 0c3082b177757d2b7e4176705e9ee53ef568a0dff838672690de4f11f2394f8e
MD5 8cfd91097c983eb8e976a2db6f416fd8
BLAKE2b-256 9d628935bcb9e3bc9447f874c52c78afa2bda6b5d12bc85e4a4067c900e0ac61

See more details on using hashes here.

File details

Details for the file memex_md-2.0.4-py3-none-any.whl.

File metadata

  • Download URL: memex_md-2.0.4-py3-none-any.whl
  • Upload date:
  • Size: 20.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.12 {"installer":{"name":"uv","version":"0.9.12"},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"22.04","id":"jammy","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for memex_md-2.0.4-py3-none-any.whl
Algorithm Hash digest
SHA256 508d84742e008708b9dc1ff5db340ee9893d1eea0221f4e8f1ba7ffe3dd199d4
MD5 5aa85873bb2121afaf73000ab54a71cc
BLAKE2b-256 3abf3b6d65bd9caa1f3763200c08c1cec56f375e08867c39c9d84bc773cd6818

See more details on using hashes here.

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