cianfhoghlaim/cianchosaint / .agents/skills/dlthub
dlthub
Verifiedname: dlthub
description: Cianfhoghlaim-side routing skill for the dltHub Platform CLI (dlthub 1.28+) — deployment manifest authoring, the dlthub run vs dlthub serve split, the 5 most common CLI errors and their recovery, and the 5-step dlthub ai status health check. Use when the user asks about "dlthub", "deploying to dlthub", "the dlthub CLI", "dlt Hub workspace", "jobs.workspace.dashboard", the dlthub run/dlthub serve verbs, or the Matched jobs are interactive error.
dlthub Platform — Cianfhoghlaim Routing Skill
This skill is the Cianfhoghlaim-side frontmatter + router for the dltHub Platform
CLI (dlthub 1.28+). The long-form runbook (5-section diagnostic tree,
canonical happy-path workflow, cross-link map) lives at
docs/agents/dlthub-run-vs-serve.md.
The 4 workbench skills below are installed into .claude/skills/ by
dlthub ai toolkit install dlthub-platform.
When to load
| User says… | Load this skill |
|---|---|
"dlthub run fails with Matched jobs are interactive" | Yes (the headline bug) |
"dlthub workspace info shows No module named 'cianfhoghlaim'" | Yes |
"dlthub ai status warns about fastmcp" | Yes |
| "How do I deploy a pipeline to dltHub?" | Yes → delegate to setup-runtime then deploy-workspace |
| "How do I serve the workspace dashboard?" | Yes → dlthub serve jobs.workspace.dashboard |
| "Where does the deployment manifest live?" | Yes → __deployment__.py |
| "What toolkits should I install?" | Yes → §"Toolkit set" below |
Decision tree → sub-skill or reference
User wants…
├── "Run a batch pipeline"
│ ├── @run.pipeline("name") in __deployment__.py ← load (deploy-workspace)
│ ├── dlthub run <script.py> ← load (deploy-workspace)
│ └── dlthub local run <job> --dry-run ← load (deploy-workspace)
│
├── "Serve an interactive notebook / dashboard"
│ ├── @run.interactive OR import marimo_module ← load (prepare-deployment)
│ └── dlthub serve <name> ← load (deploy-workspace)
│
├── "Diagnose a CLI error"
│ ├── Interactive-matched error ← runbook §2.1
│ ├── No module named 'cianfhoghlaim' ← runbook §2.2
│ ├── Workspace not yet initialized ← runbook §2.3
│ ├── FastMCP support not installed ← runbook §2.4
│ └── No toolkit with workflow installed ← runbook §2.5
│
└── "Set up / verify the workspace"
├── First-time init ← load (setup-runtime)
├── Production credentials + destination ← load (prepare-deployment)
└── Job failed / flaky ← load (debug-deployment)
Toolkit set (the 8 production toolkits)
Per the runbook §4 + the dlthub-platform skill, the 8 production toolkits to install for a full-stack setup are:
dlthub ai init --agent claude # installs `init` (rules + dlt-workspace MCP)
for t in rest-api-pipeline sql-database-pipeline \
filesystem-pipeline dlthub-platform \
data-exploration data-quality transformations; do
dlthub ai toolkit install "$t"
done
uv sync --directory cianfhoghlaim --extra dlthub-platform # installs `fastmcp-slim[server]`
claude # restart so MCP server picks up
After install, dlthub ai status should list all 8 toolkits and the only
remaining warnings should be the heuristic workspace-marker note (which
disappears once dlthub init --name cianfhoghlaim has been run from
inside the workspace root).
Cross-references
docs/agents/dlthub-run-vs-serve.md— the full diagnostic runbook (5 sections, 5 error recipes, 5-step health check, canonical happy-path)openspec/specs/dlthub-platform-integration/spec.md— the contractopenspec/changes/2026-07-06-wire-dlthub-platform-toolkits-and-deployment/— the change recorddlthub-ai-workbench/.claude-plugin/marketplace.json— the vendored workbench marketplace__deployment__.py— the deployment manifestdlt/jobs/government_circulars_job.py— the first@run.pipeline("government_circulars_ingest")job
Workbench skills (post-install)
After dlthub ai init + dlthub ai toolkit install dlthub-platform,
these 4 skills are available to the assistant:
setup-runtime— first-time workspace check (5 steps)prepare-deployment— production credentials + destination setupdeploy-workspace—dlthub deploy/dlthub local run/dlthub run/dlthub serve/ schedulingdebug-deployment—dlthub job list,dlthub job logs,dlthub info,dlthub local pipeline list
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |