From oracle-ai-data-platform-workbench-engineer-agent
Lists AIDP volumes, browses files inside, uploads/downloads via PAR flow, and creates directories. Use when staging large/binary data outside the workspace filesystem.
How this skill is triggered — by the user, by Claude, or both
Slash command
/oracle-ai-data-platform-workbench-engineer-agent:aidp-volumesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Manage AIDP volumes and their contents — the staging area for larger/binary data, distinct from the
aidp-volumes — volume files & PAR transfersManage AIDP volumes and their contents — the staging area for larger/binary data, distinct from the
workspace filesystem. Primary engine: the official Oracle aidp CLI volume group (same REST API +
auth); oci raw-request is the fallback when the CLI isn't installed.
Per references/aidp-cli-map.md: volume list | get | create | list-files | make-dir | upload-file[-with-par] | download-file[-with-par] | delete-file | delete-dir. All take
--instance-id <DATALAKE_OCID> --auth api_key --profile DEFAULT --region <r>.
# LIVE-VERIFIED: list needs BOTH catalogKey AND schemaKey
aidp volume list --catalog-key <cat> --schema-key <cat.schema> \
--instance-id <DATALAKE_OCID> --auth api_key --profile DEFAULT --region us-ashburn-1
aidp volume get <key> --instance-id <DATALAKE_OCID> --auth api_key --profile DEFAULT --region us-ashburn-1 # one volume
aidp volume list-files <key> --instance-id <DATALAKE_OCID> --auth api_key --profile DEFAULT --region us-ashburn-1 # browse files
aidp volume upload-file-with-par|download-file-with-par <key> ... # PAR transfer (binary-safe)
aidp volume make-dir|delete-file|delete-dir <key> ...
Mutating ops (create/make-dir/upload-file[-with-par]/delete-file/delete-dir): persist the body
to .aidp/payloads/ and confirm with the user before running (see references/payloads.md).
Fallback (no CLI) — same REST + auth via oci raw-request against
…/20240831/dataLakes/<DATALAKE_OCID>/… (auth ladder in references/oci-raw-request.md):
GET /volumes?catalogKey=<cat>&schemaKey=<cat.schema> (both params required; a bare path / single param
returns 400 InvalidParameter — the error names the missing key) · GET /volumes/<key> · files under
GET /volumes/<key>/files · PAR upload/download via POST /volumes/<key>/actions/… then PUT/GET the
bytes to the returned PAR URL · POST /volumes/<key>/actions/makeDir.
Verify-first (no-fabrication): the volumes routes are UNVERIFIED here —
GET /volumes?catalogKey=…alone returned 400 (also needsschemaKey). Confirm the exact params with a liveGET/volume listbefore any write/upload; record the working shape inreferences/rest-endpoint-map.md.Live-verified 2026-06-10 on de-agent — correction: the route EXISTS (not removed). A bare
GET …/volumesreturns 400 InvalidParameter naming both missing keys — "query param schemaKey must not be null; query param catalogKey must not be null" — so listing volumes requires bothcatalogKeyandschemaKeyquery params.
aidp volume list --catalog-key <cat> --schema-key <cat.schema> (read back the required-param
shape from the 400 if it errors).list-files <key>; get <key> for one volume's details.upload-file-with-par; download: download-file-with-par (PAR bytes move directly to/from
Object Storage, not through the JSON API).make-dir.Beyond file ops, the volume object has its own CRUD: aidp volume create | update | delete
(REST POST /volumes · PUT /volumes/<key> · DELETE /volumes/<key>), plus directory ops
make-dir | update-dir | delete-dir. Create body carries displayName, catalogKey, schemaKey (+
type/location) — confirm the exact fields with aidp help volume create / a live read before writing
(routes need real catalog/schema keys, like list). Grants on a volume → aidp-roles-access (volume row).
Persist create/update/delete bodies to .aidp/payloads/ and confirm first.
202 with an operation key — poll the
async-ops endpoint until terminal (see aidp-observability / oci-raw-request.md).aidp-ingest-file-to-table.aidp-ingest-file-to-table, aidp-workspace-files2plugins reuse this skill
First indexed Jun 12, 2026
npx claudepluginhub anthropics/claude-plugins-official --plugin oracle-ai-data-platform-workbench-engineer-agentManage files and notebooks in the AIDP workspace filesystem — list, upload, download, read, create, move, rename, delete. Uses the Oracle aidp CLI or OCI raw-request fallback.
Queries Databricks Unity Catalog system tables for audit, lineage, billing, compute, jobs; manages volumes with upload, download, list, and directory operations.
Provides expert guidance on Azure NetApp Files including troubleshooting, best practices, architecture, and deployment for SAP HANA/Oracle, AVS datastores, AzAcSnap, REST/PowerShell, and S3/OneLake.