eddytor-skills

Onboards new users to the Eddytor MDM platform via MCP connection, authentication, and first table creation. Activates for Eddytor onboarding, MCP setup, workspace exploration, or first table creation — even without the phrase "getting started."

Deploys self-hosted Eddytor on Azure Kubernetes Service (AKS). Builds on eddytor-deploy-helm — only the Azure-specific deltas live here. Activates for AKS, Azure Kubernetes, az aks, deploy on Azure, Azure Database for PostgreSQL Flexible Server, Azure Blob storage, or quota/vCPU errors during cluster creation.

Installs self-hosted Eddytor on any Kubernetes cluster using the published OCI Helm chart. Activates for deploying Eddytor on Kubernetes, Helm install, the eddytor chart, ingress/TLS, external vs bundled Postgres, or scaling the engine — cluster-agnostic. Cloud-specific skills (AKS/GKE/EKS/Hetzner) build on this one.

Backup, restore, and secret rotation for self-hosted Eddytor. Activates for backup, restore, disaster recovery, key rotation, rotate secrets, the encryption key (EDDYTOR_ENCRYPTION_KEY), the API key secret (EDDYTOR_API_KEY_SECRET), OAuth signing keys, Postgres backups, or recovering a lost admin key.

Deploys self-hosted Eddytor on Amazon EKS — the AWS-specific deltas on top of the cluster-agnostic Helm skill. Activates for EKS, Elastic Kubernetes Service, eksctl, deploy on AWS, RDS for Postgres, S3 object store, ALB/NLB ingress, or IRSA — even without the phrase "deploy on EKS."

Deploys self-hosted Eddytor on Hetzner Cloud — the budget path. Hetzner has no managed Kubernetes, so you bring your own k3s (hetzner-k3s or k3sup) plus the Hetzner Cloud Controller Manager and hcloud-csi driver, then install the chart. Activates for Hetzner, hcloud, hetzner-k3s, k3s on Hetzner, Hetzner Object Storage, Hetzner Load Balancer, or "cheapest way to self-host Eddytor."

Best practices for high-performance bulk reads from Eddytor over Apache Arrow Flight SQL. Activates for Flight SQL, ADBC, Arrow, bulk export, large result sets, BI tools, connecting Tableau/Power BI/pandas, or "the query is too big for REST" — even without the phrase "Flight SQL."

Best practices for the Eddytor Python SDK (`eddytor-sdk` on PyPI) — connecting, querying to pandas/Arrow, ingesting data, and using table handles. Activates for the Python SDK, eddytor-sdk, EddytorClient, query to a DataFrame, pandas/pyarrow integration, or scripting Eddytor from Python — even without "SDK."

Best practices for the `eddytor` command-line interface — auth, configuration, the verb-first command grammar, JSON output for scripting/CI, and bulk queries. Activates for the eddytor CLI, terminal usage, scripting Eddytor, CI pipelines, `eddytor login`, `eddytor get`, or `--json` — even without the word "CLI."

Best practices for calling the Eddytor REST API from apps and integrations — auth, the OpenAPI spec, the unified error envelope, rate limits, and the read/write boundary. Activates for REST API, /api/v1, HTTP integration, bearer token, API key, OpenAPI, or webhooks/automation — even without "REST."

Deploys self-hosted Eddytor on Google Kubernetes Engine. Activates for GKE, Google Kubernetes Engine, gcloud, deploy on GCP, GKE Autopilot, Cloud SQL for PostgreSQL, GCS bucket storage, Workload Identity, or a Google-managed ingress certificate — even without the phrase "deploy on GKE."

Explains who can do what in Eddytor: the four roles (Admin, Builder, Editor, Viewer), the permission matrix, minting and scoping API keys, and registering or managing OAuth clients. Activates for roles, permissions, access control, "who can do what", API key creation, OAuth client registration, audit log, or scope questions.

Installs self-hosted Eddytor on a single host with Docker Compose using the public prebuilt images. Activates for self-hosting on one VM, docker compose, the get.eddytor.dev installer, single-server Eddytor, or "no Kubernetes" — the simplest deployment path.

Enable and reason about Eddytor's telemetry — OpenTelemetry traces and metrics pushed over OTLP to your collector, plus structured logs. Activates for telemetry, OpenTelemetry, OTLP, metrics, tracing, logs, log level, monitoring, or wiring a collector. Push-only: there is no /metrics scrape endpoint.

Configures per-organisation Azure / Google OAuth applications so Eddytor can reach cloud storage with delegated (OAuth) credentials instead of static keys. Activates for "provider OAuth app", delegated storage access, Azure app registration, Google OAuth client, or "connect Azure/Google with OAuth instead of keys" — even without the word "provider."

Chooses and wires the object store that backs a self-hosted Eddytor deployment — bundled Garage for evaluation vs Azure Blob / S3 / GCS for production — and registers it right after install. Activates for "where does Eddytor store data", backing store, bundled Garage, S3/Blob/GCS for production, or object-store setup during deploy.

Browses and manages raw files in a registered Eddytor storage configuration — list, upload, download, delete, organise into folders, and move objects between configs. Activates for listing files, uploading/downloading objects, folders, moving data between buckets, or seeding a demo table — even without "object store."

Connects Eddytor to an object store (S3/MinIO, Azure Blob/OneLake, or Google Cloud Storage) so it can discover and manage tables there. Activates for connecting storage, registering a bucket or container, S3/Azure/GCS setup, credentials, or "where do my tables live" — even without the word "storage."

Renames, moves, and permanently deletes Eddytor tables, and resolves or shares them. Activates for rename table, move table to another storage, relocate, delete/drop a table, share link, or resolving a table by id — even without "lifecycle."

End-to-end guided path from an empty Eddytor instance to a constrained, populated, validated table. Activates for "set up from scratch", first table, onboarding a new dataset, "I have a CSV and nothing else", or "get me started" — even without naming a specific tool.

Imports CSV data into Eddytor tables with schema inference, type mapping, and post-import validation. Activates for CSV import, data loading, schema preview, raw CSV content, type detection, or delimiter handling — even without the word "import."

Enforces data quality rules in Eddytor through domain constraints, check constraints, validation, and profiling. Activates for data quality, validation, allowed values, constraints, profiling, domain setup, or anomaly detection — even without the phrase "data quality."

Sets up allowed value lists, category-subcategory hierarchies, and cross-table reference domains in Eddytor. Activates for allowed values, enums, dropdowns, hierarchies, parent-child relationships, lookups, foreign keys, or domain constraints — even without the word "domain."

Creates and modifies Eddytor tables, columns, primary keys, uniqueness constraints, and schema designs. Activates for table creation, schema design, column types, primary keys, add column, rename or move tables, or data type questions — even without the phrase "table management."

Inserts, updates, deletes, and upserts multiple rows atomically in Eddytor using merge_rows. Activates for bulk insert, batch update, upsert, merge rows, mass delete, data sync, or any multi-row operation — even without the word "bulk."

Migrates data into Eddytor from Excel, Microsoft MDS, SQL databases, and other MDM tools with schema mapping and validation. Activates for migration, moving data from external systems, replacing MDS, importing from databases, ETL, or switching to Eddytor — even without the word "migration."

Optimizes Eddytor table performance by compacting files, cleaning up storage, and diagnosing slow queries. Activates for optimization, compaction, vacuum, performance issues, slow queries, storage cleanup, or table maintenance — even without the word "optimization."

Reads, filters, sorts, aggregates, and analyzes data in Eddytor tables using query_rows, aggregate, and execute_sql. Activates for queries, filters, search, counts, sums, averages, group by, SQL, joins, or data exploration — even without the word "query."

Inspects table history, diffs versions, rolls back changes, restores to timestamps, and queries audit logs in Eddytor. Activates for version history, rollback, undo, restore, time travel, audit log, diffs, reverts, or previous versions — even without the phrase "version control."