5c24e67d60
Every integration (AI SDK, OpenAI, Mastra, VoltAgent, Claude memory) now calls v5 through `supermemory@5`. The config takes one `namespace` instead of `containerTags` or `projectId`, and `withSupermemory` takes `namespace` and `id` instead of `containerTag` and `customId`. No aliases. With no config the tools still use `sm_project_default`. Conversations are stored as one document per conversation keyed by `id` instead of `/v4/conversations`. `memoryForget` drops `reason`; forgetting by text previews with `forgetMatching` and then forgets exact matches by id. Claude memory marks its files with `metadata.source` instead of a second tag. Search keeps the 2.x defaults so results do not shift. `packages/ai-sdk` is removed: npm already deprecates it in favor of `@supermemory/tools/ai-sdk`, so its CI steps and playground wiring go too. `apps/sdk-playground` moves to the new tools API and v5 routes, including its Python server. Tested against production: add, search, profile, list and delete. 112 unit tests pass; type errors drop from 148 to 141, none new. The `supermemory` dependency pins the rc until 5.0.0 is published.
3 KiB
SDK Agent Playground
Chat with a real agent and switch which Supermemory SDK integration powers it.
Warning
This is a local, single-user development tool. It makes real API calls, stores browser-entered keys only in memory unless you opt into tab-scoped
sessionStorage, and exposes tools that can permanently delete documents. Use disposable development credentials and a test namespace; do not deploy it or point it at production data.
Integrations
| SDK | Style | What happens |
|---|---|---|
| AI SDK + middleware | automatic | withSupermemory injects context + saves chat |
| OpenAI + middleware | automatic | same, via OpenAI client wrapper |
| AI SDK + tools | explicit | model calls 7 memory tools via generateText |
| OpenAI + tools | explicit | OpenAI function-calling loop |
| Python OpenAI middleware | automatic | with_supermemory |
| Python OpenAI tools | explicit | SupermemoryTools loop |
| Python supermemory direct | manual | profile() + OpenAI + add() |
Setup
Prerequisites: Bun 1.3.6, Python 3.11+, and
uv. Portless is required only for the HTTPS
development hostname; the direct localhost commands below work without it.
From the repository root:
bun install --frozen-lockfile
cp apps/sdk-playground/.env.example apps/sdk-playground/.env.local
# Required:
# SUPERMEMORY_API_KEY=...
# OPENAI_API_KEY=...
The playground scripts build @supermemory/tools before starting, type-checking, or building the
Next.js app. Development mode also watches the workspace package.
Run
bun run --cwd apps/sdk-playground dev
Opens:
- Chat UI — https://sdk.dev.supermemory.ai via Portless
- Next.js server — http://127.0.0.1:3005
- Python server — http://127.0.0.1:8792
To run without Portless, use two terminals:
bun run --cwd apps/sdk-playground dev:next
bun run --cwd apps/sdk-playground dev:python
For a production-mode local smoke check, build first and then start. start
runs both the built Next.js app and the Python server, and remains intended for
local use only.
bun run --cwd apps/sdk-playground build
bun run --cwd apps/sdk-playground start
Try:
- "Remember that I prefer oat milk in coffee"
- "What do you know about my drink preferences?"
- "Forget that I like tea" (tools mode)
Env
| Variable | Required |
|---|---|
SUPERMEMORY_API_KEY |
yes |
OPENAI_API_KEY |
yes |
SUPERMEMORY_BASE_URL |
optional |
MODEL_NAME |
optional (default gpt-4o-mini) |
SDK_PLAYGROUND_PYTHON_URL |
optional (default http://127.0.0.1:8792) |
SDK_PLAYGROUND_PYTHON_PORT |
optional (default 8792) |
SDK_PLAYGROUND_ALLOW_ENV_KEYS |
optional; set true only when a trusted non-local hostname must use server env keys |
Server environment keys are exposed to the playground routes only on loopback
hosts and sdk.dev.supermemory.ai by default. Browser-provided keys remain
request-scoped and are never copied into process-global environment variables.