Skip to content

Tool availability

NovoMCP ships a 67-tool catalog. Which ones are actually usable depends on what you've wired up. The engine only exposes tools whose dependencies are met — no tool clutter, no "unavailable" errors on tools you can't run.

The three states

State What it means What you see
Available All the tool's dependencies (env vars, files) are present Tool appears in tools/list, responds to calls
Hidden A dependency is missing (e.g. GPU service URL not set) Tool absent from tools/list. Set the required env var to unlock.
Debug mode You set NOVOMCP_SHOW_HIDDEN_TOOLS=1 Every tool visible regardless of dependencies. Calls to unwired tools return "service unavailable".

Once you set the env var for a hidden tool's dependency, it appears in tools/list on the next initialize handshake. No restart required for MCP clients if they refresh their tool list.

v1 out-of-the-box tools (11)

Install the engine (pip install -r requirements.txt && python main_https.py), and you get:

  • RDKit-backed cheminformatics on user-supplied SMILES — calculate_properties, get_molecule_info, get_molecule_profile (basic path), batch_profile, screen_library
  • Public-API integrations (open APIs, no keys required) — search_chembl, search_clinical_trials, search_biorxiv. Transient network failures possible; these hit external services.
  • MD pre-flight (RCSB structure classification) — audit_system
  • Platform meta — get_platform_info
  • Autonomous discovery mode — run_novo_ag (agm). In v1 with no compute stack wired, it returns a setup-guide message pointing to which services unlock the full funnel plus a manual-workflow recipe using the always-available tools. Once you configure NOVOMCP_DB_HOST + ADDIE_MODELS_URL + AUTODOCK_GPU_URL + GROMACS_MD_URL, it runs the full 11-stage funnel.

No API keys, no databases, no compute services. Every one works on a laptop with just Python + internet.

What's NOT in v1 (but ships in the codebase, hidden)

  • search_similar, filter_molecules — need a molecule index. Hidden until NOVOMCP_MOLECULE_INDEX_URL is set. Roadmap: v1.1.5 similarity data connector ships a reference index server that reads a local parquet slice of the same corpus that unblocks omics — no paid service required.
  • Compliance path (check_compliance + tree tools) — needs a compliance service. Hidden until NOVOMCP_COMPLIANCE_URL is set.

A note on env-var naming

The engine treats molecule indexing and compliance as capabilities, not vendors. FAVES is one valid backend among several (open Kaggle-hosted index, self-hosted parquet, or your own service all work) — but the engine only ever sees the generic NOVOMCP_* env var. Point it at whichever backend you have.

Coming with future releases

Everything else is on the product roadmap — carved out cleanly, coming in weekly releases, each unlocking a slice of the catalog. Here's the map:

Omics tools — v1.1.x (SQLite bundle)

Requires: NOVOMCP_DB_HOST pointing at a Postgres with the omics schema loaded.

Unlocks: target_discovery, stratify_patients, validate_target.

Roadmap: ship data/omics.sqlite.gz + scripts/load_omics.py. One-command load into your Postgres.

Literature search — v1.2.x (PubMed fallback)

Requires today: PINECONE_API_KEY + a curated Pinecone index.

Roadmap: fallback to public PubMed E-utilities so OSS users get real literature search without Pinecone.

Unlocks: search_literature, search_patents.

Compute services — v1.3.x through v1.6.x

Each service is a Docker image (ghcr.io/novomcp/<service>:latest). Point the engine at wherever it's running:

Env var Unlocks Roadmap
AUTODOCK_GPU_URL dock_molecules, dock_with_strain v1.3.x — Modal walkthrough
GROMACS_MD_URL run_molecular_dynamics, generate_dynamics v1.4.x — Runpod walkthrough
OPENFOLD3_URL predict_structure, get_protein_structure, get_structure_result v1.5.x
NOVOMCP_QM_URL 8 quantum-mechanical tools (xTB / CREST / MCPB.py) v1.6.x
NOVOMCP_NNP_URL compute_energy, optimize_geometry_nnp (AIMNet2 / MACE / ANI-2x) v1.6.x
NOVOMCP_PROPERTIES_URL predict_pka, predict_solubility, predict_bde v1.6.x
NOVOMCP_NEB_URL find_transition_state v1.6.x
ADDIE_MODELS_URL predict_admet (31 ADMET predictions) v1.3.x (bundled with docking walkthrough)
NOVOEXPERT_URL predict_clinical_outcomes v1.7.x — NovoExpert-3 weights (MIT)
LEAD_OPTIMIZATION_URL lead_optimization v1.6.x
MOLMIM_OPTIMIZER_URL optimize_molecule v2.x — replacing MolMIM with an OSS generator

Funnel-persistence tools — v1.8.x

Requires: FUNNEL_BACKEND_URL pointing at an audit/credit-ledger service.

Unlocks: save_funnel_stage, save_funnel_context, save_funnel_memory, search_prior_runs, list_funnels, get_funnel_audit, get_funnel_context, get_pipeline_audit, get_credit_usage, generate_upload_url, get_file_status, list_files, list_jobs, get_job_status, cancel_job.

Roadmap: reference implementation of the backend + docs on plugging in your own via the pluggable spine (NOVO_AUDIT=custom).

Compliance service — v2.0.0

Requires: NOVOMCP_COMPLIANCE_URL pointing at any compliance service.

Unlocks: check_compliance, enriched paths of get_molecule_profile / screen_library / filter_molecules, explore_chemical_space, drill_into_cluster, vector_search, compare_candidates.

Roadmap: FAVES-as-API subscription at app.novomcp.com/faves (one valid backend; users can also wire their own).

Enterprise data connectors — post-v2.0

Hidden from OSS entirely. push_to_destination, pull_from_source (Snowflake, Databricks, BigQuery, Supabase) are hosted-only tools; no OSS user should see them in tools/list.

Materials Project — user brings API key

Requires: MP_API_KEY (free, sign up at materialsproject.org).

Unlocks: search_materials_project.

Debug mode

To see the full 67-tool catalog regardless of what's wired:

export NOVOMCP_SHOW_HIDDEN_TOOLS=1
python main_https.py

Tools that would normally be hidden show up in tools/list. Calls to unwired tools return structured "service unavailable" errors. Useful when developing / debugging or when you want to see everything at once.

Why hidden and not visible-with-errors

The alternative would be to always show the full 67 tools and return "service unavailable" when unwired ones are called. That's what NovoMCP used to do. It felt like a broken product.

Hidden-until-wired means: - Users see a working demo, not a lot of gray options. First impressions matter. - tools/list is the truth. Every tool in it works. LLM clients that pick tools autonomously (Claude, Cursor, Zed) never try something that will fail. - Progressive disclosure. As users add services, tools appear. The 13-tool starter set grows into the full catalog only when they've earned it by configuring what backs it.