mims-harvard Set up scVI as a remote tool
Validation status (2026-08-16): scVI dependencies, GB10 execution, loopback discovery, integration, and differential-expression calls passed on a bounded deterministic fixture with aligned finite output. Public publication, cross-user isolation, representative accuracy, broad concurrency, and recovery remain incomplete. Authenticated private Platform import and owner testing passed on 2026-08-16; public publication and independent-caller authorization/isolation remain untested.
Prerequisites
- Run from the ToolUniverse repository root on Linux with Python 3.12.3.
- GPU recommended for training; CPU only for tiny tests.
- Keep provider data, weights, caches, and credentials outside Git.
- Bind to loopback. A non-loopback bind requires TOOLUNIVERSE_API_TOKEN; never put it in arguments or results.
Run the standard-library contract check before downloading large dependencies:
python scripts/remote_validation/setup_skill_preflight.py --implementation scvi
After exporting provider resources, add --check-provider-env. After the
server starts, add --live to verify the exact MCP tool set without running
the model. Before sharing, add --check-connect-prereqs; this reports only
whether a key is set and never prints its value.
Create an isolated environment
python3 -m venv .venvs/scvi
. .venvs/scvi/bin/activate
python -m pip install --upgrade pip
python -m pip install -e .
python -m pip install -r src/tooluniverse/remote/scvi/requirements.txt
Package/network-dependent commands must be rerun in a clean environment before marking this skill complete.
Obtain credentials, data, and model weights
- Set TOOLUNIVERSE_REMOTE_DATA_ROOT. Input should contain counts and valid batch/group annotations; verify data rights.
Authorize once, then share with one short command
After installing dependencies, exporting the provider resources above, and installing the pinned relay SDK described under Connect below, run from the repository root. Log in only once per machine (and again after key rotation):
tu remote login
# Or import an existing protected 0600 file without sourcing it:
tu remote login --env-file /path/to/tooluniverse-service.env
Then each private share is one short command:
tu remote share scvi
By default, tu remote login requests a short-lived device code, opens the
TU Platform approval page, and polls until the signed-in user approves. No key
copy/paste is required. On a headless machine, add --no-browser and open the
printed link elsewhere. The CLI exchanges the approval for a computer-only key,
verifies /remote-servers/preflight, stores it in a local 0600 config file, and
never displays it.
The share command runs environment and TU Platform preflights, starts or reuses the exact loopback endpoint, validates discovery, and keeps the relay in the foreground until Ctrl-C. It automatically uses the reviewed Python, name, and worker count. Override them only when needed:
tu remote share scvi --name my-scvi-remote --workers 1
Use tu remote check scvi for a non-sharing readiness check and
tu remote run scvi for a local-only foreground server.
In an interactive terminal, sharing automatically starts the same browser flow
when the key is missing, expired, or revoked. A malformed or revoked explicit
TOOLUNIVERSE_SERVICE_KEY fails fast instead of being silently replaced; unset
or correct it, then run tu remote login. Non-interactive jobs also fail fast.
Use tu remote logout to remove only the local copy. Use tu remote logout --revoke to revoke the computer-only platform connection first; the server record remains offline for owner inspection.
Start and verify locally
mkdir -p caches/scvi runs/scvi
python -m tooluniverse.remote.scvi.scvi_tool
The Streamable HTTP endpoint is http://127.0.0.1:8010/mcp. In a second activated shell run:
python - <<'PY'
import asyncio
from fastmcp import Client
async def main():
async with Client("http://127.0.0.1:8010/mcp") as client:
print([tool.name for tool in await client.list_tools()])
asyncio.run(main())
PY
Confirm discovery contains run_scvi_integration; stop on empty, duplicate, or schema-drifted discovery.
Connect to ToolUniverse Connect
The tuplatform-connect relay is not yet published on PyPI. Install the reviewed public wheel below; its SHA-256 is pinned. Interactive sharing uses browser device authorization, so no key copy/paste or GitHub access is required.
python -m pip install fastmcp pyyaml "tuplatform-connect @ https://connect.aiscientist.tools/downloads/tuplatform_connect-0.3.0-py3-none-any.whl#sha256=3fad5eee5ecf7887a693d93ccd1aa112dc0955617a885d1fc3daded0030f9ae0"
tu doctor --forward http://127.0.0.1:8010/mcp --json
tu serve --share --forward http://127.0.0.1:8010/mcp --name validation-scvi --workers 1
Prefer browser device authorization. For CI or migration, supply
TOOLUNIVERSE_SERVICE_KEY only through a protected environment or use
tu remote login --manual-key; never put a key in shell arguments.
The authenticated 2026-08-16 Platform matrix found all 30 private owner relays online and all 41 operations discoverable. All imports remained unpublished owner drafts and were invoked through /expert-sessions/{id}/test. This implementation's draft(s) used a 120-second timeout and remote max concurrency 1.
Across the set, 38 unique operations passed return-schema and semantic validation; the three USPTO operations returned exact provider HTTP 403 and remain credential-blocked. Public publication, independent-caller authorization/isolation, broad saturation, and persistent supervision were not tested.
Run a verified example
Operation: run_scvi_integration
{"adata_path":"tiny.h5ad","n_latent":2,"max_epochs":1}
Invoke the example through the live local MCP endpoint:
python - <<'PY'
import asyncio
import json
from fastmcp import Client
async def main():
arguments = json.loads('''{"adata_path":"tiny.h5ad","n_latent":2,"max_epochs":1}''')
async with Client("http://127.0.0.1:8010/mcp") as client:
result = await client.call_tool("run_scvi_integration", arguments)
print(result)
asyncio.run(main())
PY
Expected success shape: model metadata and a finite latent representation or bounded summary. Check scientific meaning, finite values, output bounds, invalid-input behavior, and absence of paths, secrets, and traces. The success path is source-checked but not runtime-verified here unless the status note explicitly says otherwise.
Tune GPU and concurrency
- Use one worker as a conservative, unmeasured default.
- Measure cold start, two warm calls, then parallel levels 1, 2, 4, 8, and only 16 if memory permits.
- Record successes/errors, p50/p95, peak RAM/VRAM, utilization, queueing, cancellation cleanup, and recovery.
- Increase workers only after single-flight initialization and sanitized recoverable OOM/timeout behavior are proven.
Troubleshoot and clean up
- Import/executable failure: reactivate the isolated environment and reinstall its requirements.
- Missing artifact: inspect provider-only environment variables and approved relative files; never accept arbitrary caller model paths.
- 401/403 on deliberate network binding: configure matching TOOLUNIVERSE_API_TOKEN bearer auth; prefer loopback plus relay.
- Stop server/relay with Ctrl-C. If installed, run
tuplatform-service uninstall --name validation-scvi. - Revoke temporary keys. After confirmation, remove only .venvs/scvi, caches/scvi, and runs/scvi; never use a broad recursive target.
Use only official upstream documentation linked by the implementation README; do not substitute third-party model mirrors.