メニュー
Playbook for anything on the TrueNAS host — consult before touching services, data, or containers. Covers SSH entry, container interaction, and data/layout notes so you can operate safely on TrueNAS.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | truenas-docker-ops |
| description | Playbook for anything on the TrueNAS host — consult before touching services, data, or containers. Covers SSH entry, container interaction, and data/layout notes so you can operate safely on TrueNAS. |
ssh truenas (auth pre-configured)ansiblonomicon repo folder (in ansible/stacks/)/mnt/performance/docker/stacks/<stack>/compose.yaml/mnt/performance/docker/<stack>//mnt/capacity/watch/<container>/ansible/inventory/group_vars/truenas.yml in ansiblonomicon| Task | Command |
|---|---|
| List containers | ssh truenas docker ps |
| Filter containers | ssh truenas docker ps | grep <name> |
| View compose | ssh truenas cat /mnt/performance/docker/stacks/<stack>/compose.yaml |
| View logs | ssh truenas docker logs <container> |
| Exec command | ssh truenas docker exec -i <container> <command> |
| Inspect | ssh truenas docker inspect <container> |
Stacks are deployed via Ansible, not manually:
# From ansiblonomicon repo
poe truenas # Full playbook
poe truenas -t stacks # Just stacks
but you can run ad-hoc commands against the docker containers/compose files on the truenas host. Changes will not be preserved unless they are made to the ansiblonomicon repo.
/mnt/capacity/watch/anypod/data (binds to /data in container)
media/<feed_id>/transcripts/<feed_id>//data/db/anypod.dbfeed (PK id; feed metadata, source info, timestamps, counters)download (PK feed_id, id; per-episode info, status, transcript fields)appstate (PK id; last_yt_dlp_update, other global state)scripts/docker_exec_python.sh <container> '<python_code>'
scripts/docker_exec_python.sh anypod '
import sqlite3, json
conn = sqlite3.connect("/data/db/anypod.db")
conn.row_factory = sqlite3.Row
rows = conn.execute("SELECT * FROM feed").fetchall()
print(json.dumps([dict(r) for r in rows], indent=2))
conn.close()
'
Note: Default interpreter is
/app/.venv/bin/python(works for anypod). Override withCONTAINER_PYTHON:CONTAINER_PYTHON=/usr/bin/python3 scripts/docker_exec_python.sh your-container 'print("hi")'
scripts/docker_exec_sqlite.sh <container> <db_path> '<sql_query>'
scripts/docker_exec_sqlite.sh anypod /data/db/anypod.db '
SELECT feed_id, status, COUNT(*) AS count
FROM download
GROUP BY feed_id, status;
'
docker exec: Simple commands, logs, inspection# Is anypod running?
ssh truenas docker ps | grep anypod
# View anypod compose
ssh truenas cat /mnt/performance/docker/stacks/anypod/compose.yaml
# Query anypod DB
scripts/docker_exec_sqlite.sh anypod /data/db/anypod.db "SELECT * FROM feed;"
Grilling and design session that challenges your plan against the existing codebase, domain model, and documented decisions; sharpens terminology; updates documentation (CONTEXT.md, design docs) inline as decisions crystallize; and produces committable implementation plans. Use when user wants to stress-test, design, or plan a feature against their project's language and documented decisions.
Grilling and design session that challenges your plan against the existing codebase, domain model, and documented decisions; sharpens terminology; updates documentation (CONTEXT.md, design docs) inline as decisions crystallize; and produces committable implementation plans. Use when user wants to stress-test, design, or plan a feature against their project's language and documented decisions.
Capture polished screenshots of terminal applications and TUIs. Use before terminal-control when the user wants a terminal screenshot or live TUI state documented.
Plan a huge chunk of work — more than one agent session can hold — as a shared map of investigation tickets, and resolve them one at a time until the way to the destination is clear.
Review local docs and create a git commit or pull request. Use when committing changes with documentation review, or opening a PR.
Generate descriptive commit messages by analyzing git diffs. Use when the user asks for help commiting code, writing commit messages, or reviewing staged changes.