Skip to main content

Local-first durable checkpoint and dbt lineage context for coding agents

Project description

Mnemo Memory

Mnemo Memory is a local-first context service for coding agents. It stores explicit, scope-bound checkpoints and verified dbt manifest lineage in a local SQLite database, then serves a bounded, provenance-bearing context packet over MCP. It does not proxy model traffic, capture transcripts automatically, execute dbt, contact a warehouse, or call a model.

Installation

The permanent PyPI distribution name is pending maintainer approval. Once published, install the approved distribution with:

uv tool install <approved-distribution-name>
mnemo-memory --help

The installed command is deliberately mnemo-memory. It is separate from, and never replaces, an existing mnemo executable.

For a locally built wheel:

uv tool install dist/<distribution>-0.1.0a1-py3-none-any.whl

For development:

uv sync --locked
uv run mnemo-memory --help

Upgrade and uninstall after publication use the approved distribution name:

uv tool upgrade <approved-distribution-name>
uv tool uninstall <approved-distribution-name>

Quickstart

mnemo-memory init
mnemo-memory dbt ingest target/manifest.json \
  --owner-id 11111111-1111-4111-8111-111111111111 \
  --workspace-id 22222222-2222-4222-8222-222222222222 \
  --project-id 33333333-3333-4333-8333-333333333333
mnemo-memory dbt status \
  --owner-id 11111111-1111-4111-8111-111111111111 \
  --workspace-id 22222222-2222-4222-8222-222222222222 \
  --project-id 33333333-3333-4333-8333-333333333333
mnemo-memory connect codex
mnemo-memory connect claude-code

Use --data-dir for an explicit isolated local store. Otherwise Mnemo applies its documented data-directory precedence. The Codex and Claude Code connections register the absolute installed mnemo-memory launcher; they do not alter model, provider, authentication, permissions, or network settings.

See the implementation status, local MCP guide, and dbt manifest guide for lifecycle, provenance, staleness, token-budget, and recovery details.

Development verification

npm ci
npm run check
uv build --no-sources

The package requires Python 3.12.11 in the currently verified local environment. Packaging verification is performed from a temporary working directory, so runtime resources are loaded from the installed mnemo_memory package rather than from this source checkout.

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

mnemo_unified_context-0.1.0a1.tar.gz (58.0 kB view details)

Uploaded Source

Built Distribution

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

mnemo_unified_context-0.1.0a1-py3-none-any.whl (84.3 kB view details)

Uploaded Python 3

File details

Details for the file mnemo_unified_context-0.1.0a1.tar.gz.

File metadata

File hashes

Hashes for mnemo_unified_context-0.1.0a1.tar.gz
Algorithm Hash digest
SHA256 9497eb43ea783bd919179d4a1dee7502e51232396d7235db606f7ebab974a59e
MD5 940948bc081bce96dab63901b89d0279
BLAKE2b-256 491aca208a4086eab60255c2164a58991282def936e40747fa497c2e27deb98e

See more details on using hashes here.

File details

Details for the file mnemo_unified_context-0.1.0a1-py3-none-any.whl.

File metadata

File hashes

Hashes for mnemo_unified_context-0.1.0a1-py3-none-any.whl
Algorithm Hash digest
SHA256 f9b1bfc1a4b6d1652754cd299b0662119eba5fc25f5a2c65fcf9508bd3ff5a55
MD5 59e5155ab572dae87a0cc55ebe3869de
BLAKE2b-256 1a96cb2e7abaf8f161fd623eb57ac6a73a4d22ec9d944da2d0360eade654c1fb

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