Memoria
Local memory for long-running agents.
把零散上下文沉淀成可检索、可审计、可持续更新的工作记忆。
Memoria 是一个本地优先、由 LLM 驱动的工作记忆归档器。它把原始笔记、任务上下文、项目决策和 Agent 工作记录整理成 Memory Issue、Memory Chain、Proposal 和可审计的 Patch。
当前版本以 CLI 为主,数据默认保存在本机 SQLite 和 JSONL transcript 中;sleep 整理流程可以离线 mock,也可以调用 OpenAI Responses API。
安装
推荐使用 uv tool install。uv 会自动创建隔离环境,不绑定当前项目目录或 shell 里的虚拟环境:
uv tool install memoria-cli
memoria --help
也可以使用 pip;这种方式要求当前 Python 环境已经是 3.11 及以上:
python -m pip install memoria-cli
memoria --help
或者从 GitHub Releases 下载单文件二进制,不需要预装 Python。
快速开始
memoria ingest text "今天决定:Memoria MVP 先做本地 CLI,不做 Web UI。" \
--title "Memoria MVP direction" \
--tag memoria \
--hint "整理成项目决策"
memoria sleep --mock
memoria issue list --json
使用真实 LLM 整理:
export OPENAI_API_KEY="..."
memoria sleep --limit 20 --strictness balanced
文档
完整文档入口见 docs/README.md。
| 中文 | English |
|---|---|
| 安装 | Installation |
| 快速开始 | Quick Start |
| 配置 | Configuration |
| CLI 使用说明 | CLI Guide |
| 设计边界 | Design Notes |
| 开发者说明 | Development |
特性
- Local-first storage with SQLite, JSONL transcripts, zip backups, and local Git backups.
- Raw text, file, and stdin ingestion.
- LLM sleep consolidation with deterministic mock mode and OpenAI mode.
- Tool-assisted memory lookup during LLM consolidation.
- Transactional
MemoryPatchapplication with patch audit records. - Query commands for Memory Issue, Chain, Proposal, Patch, and Sleep jobs.
- Optional single-file binaries for Linux, macOS, and Windows.
状态
Memoria is early alpha software. The current scope is intentionally narrow: local CLI first, no Web UI, no API server, no MCP server, no cloud sync, and no destructive memory deletion.
Memoria 仍处于 early alpha 阶段。当前范围刻意保持克制:本地 CLI 优先,没有 Web UI、API server、MCP server、云同步,也没有破坏性删除记忆。
开发
python -m pip install -e ".[dev]"
pytest -q
uv build
See Development or 开发者说明 for the full workflow.
Metadata
Release files for memoria-cli 0.1.5
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| memoria_cli-0.1.5.tar.gz | 38.0 kB | With provenance | Details for memoria_cli-0.1.5.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| memoria_cli-0.1.5-py3-none-any.whl | Python 3 | none | any | Details for memoria_cli-0.1.5-py3-none-any.whl |
Total release size: 74.0 kB
Release files / memoria_cli-0.1.5.tar.gz
| Download URL | memoria_cli-0.1.5.tar.gz |
|---|---|
| Size | 38.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
47acf624c51f9a0ce9cc58b2b4f9b3e98f582b6b138df48c0289a00a8d57b394
|
|
BLAKE2b-256 checksum How to use checksums |
54c3d7d680fdb7dea0a779b68601f716e050e28ac3d99e81ecede917d6fb60c8
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on May 15, 2026.
Transparency logRelease files / memoria_cli-0.1.5-py3-none-any.whl
| Download URL | memoria_cli-0.1.5-py3-none-any.whl |
|---|---|
| Size | 36.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
8ae5b4d8417228e04b2637a4f877cf5dd9db97ec6aa6bdc3746eb249f5a725de
|
|
BLAKE2b-256 checksum How to use checksums |
c8ab5514cf79e106b8f8363933282398d223a7a1128ad7151c29a6313e417518
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on May 15, 2026.
Transparency logSecurity for memoria-cli 0.1.5
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
Both files in this release have been uploaded with provenance metadata.
Attestation for memoria_cli-0.1.5.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on May 15, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.