Index of every page on this site: llms.txt
All skills
k-dense-ai avatar

/benchling-integration

@ec59a29
Run it yourself

Benchling Python SDK and REST API integration for registry entities, inventory, ELN entries, workflows, Benchling Apps, and Data Warehouse queries. Use when automating lab data with benchling-sdk or the v2 API.

Use this Skill: https://skilld.dev/gh/k-dense-ai/scientific-agent-skills/benchling-integration

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesapi_endpoints.md

≈3.5k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Benchling REST API Reference

Reviewed 2026-09-30 against the official reference, platform guides, and generated benchling-api-client 2.0.434, resolved alongside SDK 1.25.0. The public reference is a JavaScript application; the published Python client supplied the endpoint/parameter/model definitions for this audit. Check your tenant's interactive reference before executing writes, especially on a validated/restricted release. No authenticated endpoint calls were made. Payloads below are illustrative.

Version and authentication

This reference covers https://{tenant}.benchling.com/api/v2. V3 also exists; consult its availability/stability guide and tenant documentation rather than replacing /v2 with /v3 in these paths.

Use HTTPS and either Basic auth (API key as username, empty password) or OAuth Bearer auth. JSON writes use Content-Type: application/json. Token exchanges are form encoded at /oauth/token; the SDK's legacy /api/v2/token path still works. See authentication for identity, expiry, and token acquisition.

Responses, pagination, and filters

Single-resource endpoints generally return a resource object; GET /entries/{id} is an exception with an entry wrapper. Creation usually returns HTTP 201 and the created resource. List responses use a resource-specific collection key, never a universal results key:

{
  "dnaSequences": [{"id": "seq_example", "name": "Construct 001", "bases": "ATCG"}],
  "nextToken": "opaque-token"
}

Most paginated lists accept pageSize (default 50, maximum 100) and nextToken. Preserve filters when requesting the next page; stop on an empty token. Some lists differ: GET /registries returns registries without the standard pagination arguments. GET /events has additional polling semantics described below.

Filters are endpoint-specific. On DNA sequences, name matches the full name; nameIncludes supports substring search. modifiedAt and createdAt are string filters; consult the tenant reference for their comparison syntax rather than assuming every timestamp filter is inclusive. Use archiveReason for archival filtering according to the endpoint definition; archived=false is not a supported replacement.

Collection and resource routes

All paths in this table are relative to /api/v2. A GET collection returns the named collection; POST creates one resource with the type's create model. A resource GET returns one object (with the entry wrapper noted above) and PATCH applies its update model.

Collection Collection methods Collection key Resource path and methods
/dna-sequences GET, POST dnaSequences /dna-sequences/{dnaSequenceId} GET, PATCH
/rna-sequences GET, POST rnaSequences /rna-sequences/{rnaSequenceId} GET, PATCH
/aa-sequences GET, POST aaSequences /aa-sequences/{aaSequenceId} GET, PATCH
/custom-entities GET, POST customEntities /custom-entities/{customEntityId} GET, PATCH
/mixtures GET, POST mixtures /mixtures/{mixtureId} GET, PATCH
/containers GET, POST containers /containers/{containerId} GET, PATCH
/boxes GET, POST boxes /boxes/{boxId} GET, PATCH
/locations GET, POST locations /locations/{locationId} GET, PATCH
/plates GET, POST plates /plates/{plateId} GET, PATCH
/entries GET, POST entries /entries/{entryId} GET, PATCH
/workflow-tasks GET, POST workflowTasks /workflow-tasks/{workflowTaskId} GET, PATCH
/folders GET, POST folders /folders/{folderId} GET
/projects GET projects /projects/{projectId} GET
/users GET users /users/{userId} GET
/teams GET teams /teams/{teamId} GET
/registries GET registries /registries/{registryId} GET

This is a task-oriented subset, not a claim that these are each service's only methods. There is no documented /users/me endpoint; test auth against an accessible resource.

Schema routes

Use GET on each collection or append /{schemaId} for one schema:

Collection Collection key
/entity-schemas entitySchemas
/entry-schemas entrySchemas
/container-schemas containerSchemas
/box-schemas boxSchemas
/location-schemas locationSchemas
/plate-schemas plateSchemas
/workflow-task-schemas workflowTaskSchemas

There is no generic stable-v2 /schemas?entityType=... route. The SDK exposes these as benchling.schemas.list_entity_schemas(), get_entity_schema_by_id(id), and analogous resource-specific methods. Schema definitions determine field names, value types, required values, units, and valid workflow statuses.

Entity create/update/archive

A registered DNA sequence create body for POST /dna-sequences:

{
  "name": "Construct 001",
  "bases": "ATCGATCG",
  "isCircular": true,
  "folderId": "lib_example",
  "schemaId": "ts_example",
  "registryId": "src_example",
  "namingStrategy": "NEW_IDS",
  "fields": {"gene_name": {"value": "GFP"}}
}

Omit registration arguments to create an unregistered entity. To assign a human registry identifier explicitly, supply entityRegistryId instead of namingStrategy, together with registryId. These are different concepts, not interchangeable IDs.

POST /rna-sequences uses bases (RNA alphabet) and isCircular. POST /aa-sequences uses aminoAcids. POST /custom-entities uses name, schemaId, folderId, and applicable fields. Do not coerce numeric fields to strings globally. A partial update uses the same field-value envelope, for example:

{"fields": {"passage_number": {"value": 16}}}

Sequence filters include folderId, schemaId, name, nameIncludes, modifiedAt, createdAt, registryId, and archiveReason; schemaId is an optional custom-entity filter rather than a requirement to list all accessible entities.

Archival endpoints are POST /dna-sequences:archive, /rna-sequences:archive, /aa-sequences:archive, /custom-entities:archive, and /mixtures:archive. Use the appropriate array key (dnaSequenceIds, rnaSequenceIds, aaSequenceIds, customEntityIds, or mixtureIds) and an allowed reason:

{"dnaSequenceIds": ["seq_example"], "reason": "Retired"}

Allowed EntityArchiveReason values in this release include Made in error, Retired, Expended, Shipped, Contaminated, Expired, Missing, and Other. Free-form reasons such as Cleanup are not the enum. Read the archival response rather than assuming deletion; archival preserves records and can affect related data.

For mixture creation, each ingredient uses a separate amount and unit:

{
  "name": "Example formulation",
  "schemaId": "ts_mixture",
  "folderId": "lib_example",
  "ingredients": [{
    "componentEntityId": "bfi_component",
    "amount": "100",
    "units": "mg",
    "catalogIdentifier": null,
    "componentLotContainerId": null,
    "componentLotEntityId": null,
    "componentLotText": null,
    "notes": null
  }]
}

Inventory payloads and actions

For container/box/plate creation, schemaId is required. Location creation requires name and schemaId. Depending on storage type, add parentStorageId, barcode, name, or schema fields. Query containers and boxes by ancestorStorageId and barcodes (comma-separated); parentStorageId and singular barcode are not the list filters shown in this SDK. Ancestor filters include nested descendants.

Move a container physically with PATCH /containers/{containerId}:

{"parentStorageId": "loc_destination"}

Check out with POST /containers:check-out:

{"containerIds": ["con_example"], "assigneeId": "usr_example", "comment": "At bench"}

Check in with POST /containers:check-in:

{"containerIds": ["con_example"], "comments": "Returned"}

These operations return HTTP 200 with an empty JSON object {} on success. There is no locationId parameter for check-in. Notice the hyphen in each route and different comment/comments keys.

Material transfer has two distinct operations:

  • POST /containers/{destinationContainerId}:transfer: ContainerTransfer body, including destinationContents plus the applicable source/quantity fields; HTTP 200 with {} on success.
  • POST /transfers: a transfers array of MultipleContainersTransfer objects; returns HTTP 202 with a taskId for async processing. Limit: 5000 transfers per request.

Illustrative batch body:

{
  "transfers": [{
    "sourceContainerId": "con_source",
    "destinationContainerId": "con_destination",
    "transferQuantity": {"value": 5.0, "units": "uL"}
  }]
}

Neither /containers:transfer nor /containers:bulk-transfer is the corresponding stable-v2 route. Never use material transfer to represent moving a tube to a box.

Plate creation wells is a mapping of positions to well properties, not an array of {"position": "A1", "entityId": ...}. Create/read the plate first, resolve its well container IDs, then load biological contents through transfer APIs. Storage schema and well positions must match the actual plate configuration.

Entries, workflow tasks, folders

POST /entries requires name and folderId; entryTemplateId, initialTables, schemaId, and fields support appropriate template/schema workflows. PATCH /entries/{id} updates supported metadata and schema fields; it is not a general rich-text editor. GET /entries supports projectId, schemaId, modifiedAt, and other documented filters, but does not expose folderId in the stable SDK's list method. GET /entry-templates and /entry-templates/{entryTemplateId} discover templates. There is no general entry_links service to attach arbitrary objects to an entry.

POST /workflow-tasks requires a task group:

{"workflowTaskGroupId": "wtg_example", "assigneeId": "usr_example"}

GET /workflow-tasks accepts comma-separated workflowTaskGroupIds, statusIds, and assigneeIds, plus schemaId and pagination. PATCH /workflow-tasks/{workflowTaskId} can set statusId, assigneeId, scheduledOn, and fields. Status IDs come from the workflow task schema; there is no workflowId create/list field here.

POST /folders takes name and parentFolderId; the parent determines the project. Do not add a projectId to FolderCreate. Folder listing does accept projectId and parentFolderId filters.

Async jobs and events

GET /tasks/{taskId} polls an asynchronous server job; it is not the workflow-task CRUD endpoint. Responses contain status (RUNNING, SUCCEEDED, FAILED) and operation-specific response, errors, or message when applicable. Timeout on the client does not cancel the job. Keep the task ID and check its final outcome.

GET /events returns events with nextToken. Filters include createdAt.gte, startingAfter (event ID), eventTypes (comma-separated), poll, and pageSize. The SDK spelling is created_atgte. Events are sorted by processing order, not by createdAt; never terminate recovery early merely because a timestamp is out of order. See event recovery for retention and deduplication.

Field values

The fields mapping is keyed by actual schema field names, with a value member:

{
  "passage_number": {"value": 15},
  "comment": {"value": "QC passed"},
  "linked_entities": {"value": ["seq_example"]},
  "selection": {"value": ["sfso_example"]}
}

Use schema-defined scalar/list multiplicity. Dropdown values are option IDs, not human option labels. Numeric fields accept numeric values; date/datetime formats depend on the field definition. null is distinct from omission and may clear a value. See schema field types.

Errors and limits

Failures normally use an error object with message, type, and sometimes userMessage. Do not assume exception-class names like RateLimitError appear in JSON. Check HTTP status before reading success data: 400 invalid request, 401 authentication, 403 permission, 404 missing/inaccessible resource, 429 throttling, and 5xx server errors. A successful check-in or single-container transfer returns an empty JSON object.

The V2 limits guide currently lists 60 requests/30 seconds across user keys per tenant, 300/30 seconds per app, and 1000/30 seconds across apps per tenant. There are separate dynamic throughput limits. The response headers are x-rate-limit-limit, x-rate-limit-remaining, and x-rate-limit-reset; reset is seconds remaining, not an epoch timestamp.

Use bounded exponential backoff with jitter for 429/transient server failures and honor server retry guidance when present. Do not invent a JSON retryAfter field or recurse without an attempt limit. Reconcile ambiguous write failures before repeating creates.

Direct pagination example

This illustrative read-only example uses httpx, checks status, and preserves filters:

import os
import httpx

with httpx.Client(
    base_url=os.environ["BENCHLING_TENANT_URL"].rstrip("/") + "/api/v2/",
    auth=(os.environ["BENCHLING_API_KEY"], ""),
    timeout=30.0,
) as client:
    params = {"pageSize": 100, "schemaId": "ts_example"}
    while True:
        response = client.get("dna-sequences", params=params)
        response.raise_for_status()
        data = response.json()
        for sequence in data["dnaSequences"]:
            print(sequence["id"], sequence["name"])
        next_token = data.get("nextToken")
        if not next_token:
            break
        params["nextToken"] = next_token

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No alerts27d4 checks · Risk SAFE
  • Gen Agent Trust Hub27d

    This skill provides a secure framework for integrating with the Benchling platform. It correctly handles authentication via environment variables, uses pinned versions for its dependencies, and restricts network activity to the official Benchling API and trusted academic sources. No malicious patterns or security risks were identified.

  • Socket1mo

    No alerts

  • Snyk1mo

    Risk: LOW · No issues

  • ZeroLeaks7mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at ec59a29. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 4 days ago.

Activeupdated 2 months ago
All 1 allowed tools
Read Write Edit Bash
Other metadata
compatibility
Requires Python 3.9+, benchling-sdk 1.25.0, network access, a Benchling tenant with API access, and API key or OAuth app credentials.
metadata
{
  "version": "1.7",
  "last-reviewed": "2026-09-30",
  "skill-author": "K-Dense Inc.",
  "openclaw": {
    "primaryEnv": "BENCHLING_API_KEY",
    "envVars": [
      {
        "name": "BENCHLING_TENANT_URL",
        "required": true,
        "description": "Benchling tenant base URL."
      },
      {
        "name": "BENCHLING_API_KEY",
        "required": false,
        "description": "API key auth (alternative to OAuth)."
      },
      {
        "name": "BENCHLING_CLIENT_ID",
        "required": false,
        "description": "OAuth app client id."
      },
      {
        "name": "BENCHLING_CLIENT_SECRET",
        "required": false,
        "description": "OAuth app client secret."
      },
      {
        "name": "BENCHLING_PROD_TENANT_URL",
        "required": false,
        "description": "Production tenant URL (multi-env setups)."
      },
      {
        "name": "BENCHLING_PROD_API_KEY",
        "required": false,
        "description": "Production API key (multi-env setups)."
      },
      {
        "name": "BENCHLING_STAGING_TENANT_URL",
        "required": false,
        "description": "Staging tenant URL (multi-env setups)."
      },
      {
        "name": "BENCHLING_STAGING_API_KEY",
        "required": false,
        "description": "Staging API key (multi-env setups)."
      }
    ]
  }
}

README badge

README badge for k-dense-ai/scientific-agent-skills/benchling-integration