supermemory_profile_search is async, but without the optional aiohttp
extra it falls back to requests.post, which blocks whatever event loop
awaited it for the whole request. aiohttp is not installed by default,
so on a plain pip install every memory search on the async path stalls
the caller's loop.
The fallback now runs requests.post through asyncio.to_thread. The new
test forces the fallback and fails if the request holds the loop.
## Stack Context
Part 2 of a 3-PR stack moving memory deduplication into the SDKs. See `sdk-dedup/tools-ts` (parent) for the full context and the TypeScript implementation this mirrors.
## What?
Port the normalized, priority-ordered (`static > dynamic > search`) profile deduplication into the Python SDKs.
- Each request injects one **owned memory block that replaces** the prior block rather than accumulating.
- Dedup is **request-local** (no shared state), so it stays correct under concurrency.
Covers OpenAI, Agent Framework (middleware + context provider), Cartesia, and Pipecat.
## Why?
Keeps the Python SDKs at behavioral parity with the TypeScript SDK so all integrations deduplicate memory the same way.
## Testing
- OpenAI: 31 passed, 11 skipped (live)
- Agent Framework: 59 passed
- Cartesia: 8 passed
- Pipecat: 8 passed
🤖 Generated with [Claude Code](https://claude.com/claude-code)
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> **Medium Risk**
> Changes memory formatting and system-prompt injection across multiple SDK integrations; incorrect dedup or replacement could alter LLM context, but there is no auth or data-store risk.
>
> **Overview**
> Ports **normalized cross-source memory deduplication** and **replace-not-append injection** into the Python OpenAI, Agent Framework, Cartesia, and Pipecat packages so they match the TypeScript SDK behavior.
>
> **Deduplication** uses request-local keys: strip optional `[YYYY-MM-DD]` prefixes, normalize whitespace, and compare with `casefold`, with priority **static → dynamic → search**. In **`query` mode**, profile static/dynamic are excluded from dedup input so facts that only appear in search (or overlap profile) are not dropped before formatting.
>
> **Injection** no longer appends memory text every turn. OpenAI and Agent Framework middleware **strip prior owned `<supermemory context="user-memories" readonly>` blocks** and **replace** them once per request while keeping the caller’s system instructions; extra system messages lose stale blocks only. New helpers (`strip`/`replace`/`wrap`) live in each package’s utils.
>
> Tests cover normalized fact variants, query-mode search retention, and stale block replacement.
>
> <sup>Reviewed by [Cursor Bugbot](https://cursor.com/bugbot) for commit 42f308b224. Bugbot is set up for automated code reviews on this repo. Configure [here](https://www.cursor.com/dashboard/bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
## Summary
- Expand `SupermemoryTools` from 2 tools to 7 (matches `@supermemory/tools`)
- Add `memory_forget` via shared HTTP helper
- Add document list/add/delete and get_profile tool surfaces
- Expand tests for new tools and execution paths
Stacked on #1429
## Test plan
- [x] `uv run pytest tests/test_tools.py::TestMemoryOperationsUnit`
Made with [Cursor](https://cursor.com)
## Summary
- Replace deprecated `search.execute` with `search.memories` (hybrid mode) in `search_memories`
- Replace `memories.add` with `client.add` in tools and middleware
- Fix middleware `container_tag` param (was incorrectly `container_tags`)
- Fix profile memory deduplication for string and Pydantic API items
- Bump `supermemory>=3.50` and `requires-python>=3.9`
## Test plan
- [x] `uv run pytest tests/test_tools.py::TestMemoryOperationsUnit`
Made with [Cursor](https://cursor.com)
