Files
mem0/integrations/mem0-strands/README.md
Saket Aryan df70d7833f fix(plugins): stamp surface identity at record time, not send time
Six defects in 0.3.x plugin telemetry. Defects 1, 2 and 6 were not three bugs:
they were one spool protocol getting three properties wrong.

Identity was decided by the wrong process. `harness` was stamped in record(),
correctly, but `source` was read from a module global in flush() — so whichever
process drained the spool named every event in it. Two processes never call
init(): mcp_server.py, and the detached `python3 telemetry.py` sender that
spawn_flush() starts. record() now stamps source beside harness, and the build
generates core/_harness_id.py per host so identity resolves with no init() call
at all. That also unifies two defaults that disagreed (`<host>_plugin` vs
`MEM0_<HOST>_PLUGIN`), which could yield three source values for one plugin.

Ownership was inferred, not held. Path.replace is os.rename, which preserves
mtime, so a claim made after a quiet minute inherited the spool's age and was
stealable the instant it existed. Claims are touched on creation and the
per-batch rewrite doubles as a lease heartbeat.

Progress was not durable. flush() returned on the first failed batch without
truncating, so the retry re-posted from index 0 — 150 events delivered 250
times. It now rewrites the claim with the unsent remainder after every batch,
bounding a crash to one repeated batch, and each event carries a uuid.

Parked batches starved. They were only reachable when no spool existed, and
because sessions keep recording there usually was one, so a batch parked by a
failed send waited until the 7-day expiry deleted it unsent — despite its own
presence being what starts the sender. flush() drains them in the same run, and
expiry now applies only after a genuine retry has failed.

code.install counted upgrades and repeat sessions. is_first_run() read the
identity file, which only a successful flush writes, so an offline user recorded
an install every session forever. A dedicated install-state.json is claimed
atomically at record time; a non-empty data directory reads as an upgrade.

The docs called this anonymous. Every event carries the account email, and the
hashes were unsalted SHA-256 over a git remote URL or an absolute path
containing the username. READMEs, the module docstring and a new docs section
now say what the code does, and repo/session digests are salted per install.

A cached email outlived an API key change. It is now re-resolved when the key's
fingerprint differs, and $identify aliases anonymous->email only — aliasing one
account to another merges person profiles irreversibly.

All six shipped green because the shared core's only tests lived under one host,
behind a conftest that calls init() at import. Core behaviour was never
exercised uninitialised. Adds agent-plugin-core/tests with no init, including
subprocess tests and coverage for the portable plugin, which has no flush worker
and would pass a native-only test vacuously.

Also puts the three surface headers on the SDKs, CLIs and integrations, and
corrects a README claiming ZAPIER/STRANDS were already in the platform allowlist.

Verified: 59 core tests, 203 claude-code, 11 cursor, 5 codex, 2 kimi, 6
antigravity. ruff and compileall clean. --check clean for all six hosts.
TypeScript changes are not typechecked locally (deps not installed).

Claude-Session: https://claude.ai/code/session_01C7tEmH86HAr7GoAAKCEHZb
2026-09-15 00:14:11 +05:30

104 lines
4.7 KiB
Markdown

<div align="center">
<h1>mem0-strands</h1>
<h3>Persistent long-term memory for Strands Agents, backed by Mem0</h3>
<p>
A community <a href="https://strandsagents.com/">Strands Agents</a> integration that plugs
<a href="https://mem0.ai">Mem0</a> in as a first-class <code>MemoryStore</code>.
</p>
</div>
---
`mem0-strands` gives [Strands](https://github.com/strands-agents/sdk-python) agents durable memory
that survives across sessions, backed by [Mem0](https://mem0.ai). Where the
`mem0_memory` tool is called explicitly by the model, `Mem0MemoryStore` plugs into the **agent loop**
directly: the manager recalls context and injects it automatically, and writes new memories, either
verbatim or by extracting facts from the conversation.
- **Automatic recall + injection** — relevant memories are searched and prepended to the prompt every turn, no tool call required.
- **Server-side extraction** — raw conversation turns are handed to Mem0, which distills and de-duplicates facts on its own pipeline (no extra client-side model call).
- **Hosted or self-hosted** — the managed [Mem0 Platform](https://app.mem0.ai) by default, or your own Mem0 OSS backend via a config dict.
## Install
```bash
pip install mem0-strands
```
## Usage
```python
from strands import Agent
from strands.memory import MemoryManager
from mem0_strands import Mem0MemoryStore
# Recall + write, distilling facts from the conversation via Mem0's server-side extraction.
store = Mem0MemoryStore(user_id="alex", writable=True, extraction=True)
agent = Agent(memory_manager=MemoryManager(stores=[store]))
# The agent now recalls from and writes to Mem0 without any explicit tool call.
agent("Remember that I prefer dark-mode dashboards and only drink oat milk.")
agent("How do I like my dashboards?") # recalls the stored preference
```
Set `MEM0_API_KEY` for the hosted platform (get one at [app.mem0.ai](https://app.mem0.ai)), or pass
`api_key=...`. For a self-hosted Mem0 OSS backend, pass a `config=...` dict instead.
## How it works
`Mem0MemoryStore` implements all three `MemoryStore` hooks:
| Method | Maps to | When it runs |
|---|---|---|
| `search(query)` | `mem0.search(query, filters={...})` | Every turn, to recall and inject context |
| `add(content)` | `mem0.add(content, infer=False)` | The `add_memory` tool / a client-side extractor — stores a fact verbatim |
| `add_messages(messages)` | `mem0.add(rendered_turns, infer=True)` | Extraction — renders conversation turns to text, then hands them to Mem0's **server-side** extraction |
Because `add_messages` is implemented, enabling `extraction` routes conversation turns straight to Mem0's own
extraction pipeline. A store that only implemented `add` would instead need a client-side `ModelExtractor`
(an extra model call) to distill facts first.
### Configuration
| Argument | Default | Description |
|---|---|---|
| `user_id` / `agent_id` / `run_id` / `app_id` | _(at least one required)_ | Mem0 entity scope that owns the memories |
| `name` | `"mem0"` | Store identifier, used to target it from memory tools |
| `writable` | `True` | Whether the manager may write to the store |
| `extraction` | `None` | Automatic extraction (`bool` or `ExtractionConfig`) |
| `max_search_results` | `None` | Default result cap per search (falls back to 5) |
| `metadata` | `None` | Default metadata merged into every write |
| `api_key` / `host` | env | Mem0 platform key / base URL (`api_key` defaults to `$MEM0_API_KEY`) |
| `config` | `None` | Mem0 OSS config dict for a self-hosted backend |
## The explicit tool
For the model-called tool (`store` / `retrieve` / `get` / `delete`), use the
[`mem0_memory`](https://github.com/strands-agents/tools) tool from `strands-agents-tools`. The store and
the tool share one Mem0 backend and namespace.
## Telemetry
The store sends usage events (store configuration, operation, duration, result
counts, coarse failure kind) over the Mem0 SDK's existing telemetry client,
tagged `source="STRANDS"`. These are **not anonymous**: when an API key is
configured they are sent under your Mem0 account email, the same way the SDK
attributes its own. Queries, memory text, message content, entity ids, and
metadata are never sent. Turn it off with `MEM0_TELEMETRY=false`.
## Development
The package lives under [`python/`](python/) (monorepo-style layout matching the
[Strands extension-template](https://github.com/strands-agents/extension-template)).
```bash
cd python
pip install hatch
hatch run test # pytest (no live server required — mocked client)
hatch run prepare # format + lint + typecheck + test
```
## License
[Apache-2.0](LICENSE). Mem0 is a trademark of its respective owner. Strands Agents is a project of its respective authors.