MDEngine

Measured conclusion: Alive, not deep-tested. L2 outcome: not_attempted.

MD workbench for agents: 13 analysis tools, renders, hosted GPU runs; LAMMPS + OpenMM.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "MDEngine": {
      "url": "https://api.forcefieldsilicon.com/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (14)

• guide read-only

  No key needed. Without customer_type: the list of guides (one line each), the start page and how to get a key. With customer_type (mineral-processing | dark-proteins): the full guide, which says what to supply, what runs, what comes back and what it costs. With template (openmm-stability): the script text of that template, ready to save and submit.

  Parameters:

  • template — string — openmm-stability (dark-proteins stage 1 OpenMM script); returns the script text
  • customer_type — string — mineral-processing | dark-proteins; omit for the list

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• campaign_request

  No key needed. Sends a study request to a person and returns a request_id, the tier that fits, the price band, what we still need from you, the next step and the claim scope. Nothing is run or charged. Call it after reading a guide.

  Parameters:

  • customer_type (required) — string — mineral-processing | dark-proteins | semiconductor | other
  • question (required) — string — One paragraph: the decision the study feeds
  • contact_email (required) — string — Where the reply goes
  • grid — string — Free text: compositions x temperatures, or ligands x pockets (e.g. 3 ligands x 1 receptor)
  • data_rules — string — How the data and results may be used, kept or shared
  • organization — string — Optional
  • inputs_available — array — What you can supply now, e.g. bath range, AlphaFold model, ligand list

  Hints: read-only no · destructive no · idempotent no · open-world yes

• account read-only

  Balance in USD (`balance_usd`), the part of it held by your unfinished jobs (`reserved_usd`, `held_jobs`) and what a new submit can use (`available_usd` = balance - reserved; a submit needs wall_limit_s x rate of it), how jobs are priced (`pricing.mode` job = the deck's own work in atom-steps at per-class prices, capped at wall_limit_s x rate; metered = per second of pod time), the rate table, the key id of the API key in use, and `packs` (when the endpoint sells credit): [{id, name, usd, gpu_h, url}] to top up THIS key by appending ?client_reference_id=<key_id> to the pack url.

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• submit_job

  One call: create a hosted GPU job, upload the deck given INLINE as {relative_path: text}, and queue it. Total inline size <= 8 MB; for larger decks use create_job, PUT the tarball to upload_url, then start_job. Billing starts at the first heartbeat (state running) and stops at done/failed/cancelled. The deck is executed as a program on an isolated GPU pod.

  Parameters:

  • input (required) — string — Relative path of the LAMMPS (or runner) input script inside the deck, e.g. in.lmp
  • files (required) — object — Deck contents: {"in.lmp": "...", "data.al": "..."}; paths relative, no '..'; must include `input`
  • gpu — string — GPU class; call account for the offered classes and how jobs are priced. Default any = cheapest available
  • label — string — Free text <= 120 chars shown in job lists
  • launch — string — Launch template; omit for the default KOKKOS/CUDA LAMMPS command line
  • runner — string — Runner flavour: lammps (default) or openmm (beta)
  • end_marker — string — Optional line your deck prints when it truly finished (e.g. a `print DONE` at the end); a done job whose log lacks a line starting with it gets verdict=warnings, end_marker_missing=true
  • estimate_s — integer — Your runtime guess in seconds; informational only; the pre-check uses wall_limit_s net of credit held by your unfinished jobs
  • wall_limit_s — integer — Hard cap in seconds (default 14400, max 86400); the job fails at the cap and is billed to it

  Hints: read-only no · destructive no · idempotent no · open-world yes

• create_job

  Step 1 of the two-step path for big decks: validates the spec, reserves a job id, returns a presigned upload_url. PUT the deck as a .tar.gz (<= 2 GB, relative paths, input at `input`) to upload_url, then call start_job.

  Parameters:

  • input (required) — string — Relative path of the LAMMPS (or runner) input script inside the deck, e.g. in.lmp
  • gpu — string — GPU class; call account for the offered classes and how jobs are priced. Default any = cheapest available
  • label — string — Free text <= 120 chars shown in job lists
  • launch — string — Launch template; omit for the default KOKKOS/CUDA LAMMPS command line
  • runner — string — Runner flavour: lammps (default) or openmm (beta)
  • end_marker — string — Optional line your deck prints when it truly finished (e.g. a `print DONE` at the end); a done job whose log lacks a line starting with it gets verdict=warnings, end_marker_missing=true
  • estimate_s — integer — Your runtime guess in seconds; informational only; the pre-check uses wall_limit_s net of credit held by your unfinished jobs
  • wall_limit_s — integer — Hard cap in seconds (default 14400, max 86400); the job fails at the cap and is billed to it

  Hints: read-only no · destructive no · idempotent no · open-world yes

• start_job

  Step 2: queue a job whose deck tarball has been uploaded. A GPU pod is launched; billing starts when it reports running.

  Parameters:

  • id (required) — string — Job id, e.g. MDJOB-20260907-3F200C

  Hints: read-only no · destructive no · idempotent no · open-world yes

• job_status read-only

  State (created|uploaded|queued|launching|running|uploading|done|failed|cancelled), GPU, rate, billed seconds, cost so far, exit code, error, last thermo lines, and the post-run verdict: clean | warnings (a done run whose log shows QEq CG failures, lost atoms, a fix halt, or no end_marker line; counts in `warnings`).

  Parameters:

  • id (required) — string — Job id, e.g. MDJOB-20260907-3F200C

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• job_log read-only

  The last <= 20 thermo/log lines the running pod reported (30 s heartbeat). Full log.lammps is in the results tarball.

  Parameters:

  • id (required) — string — Job id, e.g. MDJOB-20260907-3F200C

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• job_results read-only

  For a done/failed job: a presigned download_url (valid ~7 days) for the results tarball (work/, log.lammps, exitcode). Results are deleted 30 days after the run.

  Parameters:

  • id (required) — string — Job id, e.g. MDJOB-20260907-3F200C

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• list_jobs read-only

  Jobs of this API key, newest first. Compact rows (id, state, label, created, finished, gpu, cost_usd, error) unless full=true.

  Parameters:

  • full — boolean — Return the complete status object per job (default false)
  • limit — integer — 1..200, default 20

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• delete_results destructive

  For a finished job: delete its deck and results tarballs immediately instead of at the automatic 30-day purge. Metadata and billing records are kept; results can no longer be downloaded.

  Parameters:

  • id (required) — string — Job id, e.g. MDJOB-20260907-3F200C

  Hints: read-only no · destructive yes · idempotent yes · open-world yes

• capabilities read-only

  Capability manifest of the hosted runners: LAMMPS version, installed packages, and every style by category with gpu=true (KOKKOS-accelerated) or gpu=false (exists, but runs on the pod's CPU cores at the GPU rate). Default = compact summary; runner=lammps&full=true returns the whole style table. Use preflight_deck to check a specific deck.

  Parameters:

  • full — boolean — Include the full style table (large)
  • runner — string — lammps | openmm; omit for all

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• preflight_deck read-only

  Dry run of the check submit_job performs: which styles the deck asks for are MISSING on every hosted LAMMPS image (the run would exit at startup), which are CPU-only, whether its pair style will use the GPU at all, and which image (`runner`) the job will be routed to — decks needing packages beyond the fast default image run on lammps-full automatically. Nothing is created or billed.

  Parameters:

  • input (required) — string — Relative path of the LAMMPS (or runner) input script inside the deck, e.g. in.lmp
  • files (required) — object — Deck contents {path: text}, must include `input`

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• cancel_job destructive

  Cancel a job that is not finished. A running job is billed up to the cancel time; its pod is terminated.

  Parameters:

  • id (required) — string — Job id, e.g. MDJOB-20260907-3F200C

  Hints: read-only no · destructive yes · idempotent yes · open-world yes

Publisher website Repository

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.