Skip to main content

Memtru

Reproducible tests for the memory systems behind AI agents. Memtru writes synthetic markers into your memory store and checks the boundaries that matter: one user's memory never reaches another user or tenant, deleted memory stays deleted, updates replace stale facts, and sessions stay separate.

Every finding comes with the evidence to reproduce it. Tests run locally or in CI; nothing is uploaded unless you run memtru upload.

pip install memtru
memtru demo              # a deliberately broken store, then a fixed one
memtru demo --afterlife  # a deleted memory that comes back after a later event

Test your own system

memtru init              # writes memtru.yaml
memtru doctor            # checks the configuration and adapter
memtru test              # runs the scenarios, writes JSON and HTML reports

Adapters: the built-in reference store (in_memory), any service implementing the Memtru REST contract (http), your own Python class (custom), and Mem0 OSS (pip install "memtru[mem0]").

Scenarios

Id What it checks
MT-ISO-001 Cross-user isolation
MT-TEN-002 Cross-tenant isolation
MT-DEL-003 Deleted memory stops being retrievable
MT-STALE-004 Updated facts replace stale ones
MT-SES-005 Session isolation
MT-INT-006 Updates change only the intended record
MT-POISON-007 Controlled memory poisoning (agent mode)
MT-AF-001 Afterlife: a late write brings deleted memory back
MT-AF-002 Afterlife: a new session rebuilds deleted memory from history

Every scenario reports PASS, FAIL, INCONCLUSIVE, ERROR or SKIPPED. A result is never PASS without a positive control showing the authorized path works.

Exit codes make it CI-ready: 0 pass, 1 a scenario failed, 3 execution error, 4 inconclusive when --fail-on-inconclusive is set.

License

Proprietary. Free to install and use for testing systems you own or are authorized to test; see the LICENSE file for the terms. Test results describe the scenarios that ran and are not a guarantee of security.

Metadata

Release files for memtru 0.1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for memtru 0.1.0
File Size Uploaded Provenance
memtru-0.1.0.tar.gz 64.1 kB With provenance Details

Built distribution (wheel)

Table of built distributions (wheels) for memtru 0.1.0
File Interpreter ABI Platform
memtru-0.1.0-py3-none-any.whl Python 3 none any Details

Total release size: 146.9 kB

Release files / memtru-0.1.0.tar.gz

Download URL memtru-0.1.0.tar.gz
Size 64.1 kB
Tags Source
SHA-256 checksum
How to use checksums
884e843b266c7697feee40ab1d36603dce5b4739e0e610b623cb766de919ad43
BLAKE2b-256 checksum
How to use checksums
d06034aa471888ecc58c6dd1a5148dbf3db3e991eae610deb2ab5b723a7094fa
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

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 Oct 23, 2026.

Transparency log

Release files / memtru-0.1.0-py3-none-any.whl

Download URL memtru-0.1.0-py3-none-any.whl
Size 82.8 kB
Tags Python 3
SHA-256 checksum
How to use checksums
82df46d1a0594f835b5267b0b977bf884284d08be0ac45a77f043c4108b4efdf
BLAKE2b-256 checksum
How to use checksums
1a7e4fd025b496e563864c4d1c3832d43cbd1dac65f6843461ec78c2ff95b5d4
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

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 Oct 23, 2026.

Transparency log

Security for memtru 0.1.0

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

All 2 files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for memtru-0.1.0.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 Oct 23, 2026.

Transparency log

A 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.

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 release files Full provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page