redth/homelab-assist/homelab-research
Claude Code and agent skill by redth/homelab-assist
- community
Checking that the SKILL.md file is still in the repository.
How to research version-correct, canonical answers for homelab software (Proxmox, Docker, Dockhand, Portainer, nginx-proxy-manager, etc.) without being misled by stale search results. Load this whenever you are not fully confident about a configuration, command, API field, or behavior before acting on a homelab system.
Install
npx vskill@latest install redth/homelab-assist/homelab-researchSKILL.md
Researching homelab problems correctly
Homelab software moves fast and the web is full of outdated guides. A confidently-wrong answer here can cause an outage or data loss. When you are not certain, slow down and research the version-correct, canonical answer before acting.
Rule 0 — establish the version first
Before trusting any guidance, find out what the system is actually running:
- Proxmox:
pveversion(orpveversion -v), or APIGET /version. PVE 8.x vs 9.x differ in real ways. - Docker:
docker version,docker compose version. - Dockhand / Portainer / Dockge / NPM: check the app's version (UI footer,
/apiversion endpoint, image
tag). MCP/API clients are often version-matched (e.g. Portainer MCP ~=2.42 ↔ Portainer 2.42.x).
Then scope every search and every doc page to that version. A correct answer for 8.x can be wrong for 9.x.
Rule 1 — prefer canonical, AI-optimized sources
Order of trust:
- Official versioned docs for the exact version (see
reference/doc-sources.md). - `llms.txt` / AI-optimized docs when the project publishes them (e.g. Docker:
https://docs.docker.com/llms.txt). These are clean, current, token-efficient — fetch them first.
- Official API viewers / OpenAPI specs for exact field names and shapes (Proxmox API viewer; Portainer
api-docs).
- Project source / release notes / changelog for "did this change recently?" questions.
- Forums / blogs / Reddit only to find leads, then verify against the canonical source. Treat any
undated or old post as suspect — check whether it predates the running version.
The curated, maintained source list lives in `reference/doc-sources.md`. Consult it; keep it updated when you find a better source.
Rule 2 — watch for version-rot signals
Be extra skeptical when you see:
- Commands or kernel modules that may have been removed/renamed (e.g. Proxmox
vfio_virqfdis gone on 9.x). - API fields or endpoints that moved (e.g. PVE 9.x VNC proxy changes).
- "Just set X" guides with no date and no version — verify before applying.
- Advice that contradicts the official docs — the official docs win for the running version.
Rule 3 — verify on the actual system (read-only)
Confirm assumptions against the live system using read-only checks before changing anything:
- Does the field/endpoint exist on this version? (query it)
- Does the device/path/storage actually exist? (
lspci,lsblk, list storage — read-only) - What is the current config right now? (read it before editing)
If reality contradicts the doc/plan, stop and reconcile — don't push a change built on a wrong premise.
Rule 4 — when still uncertain, say so
If after researching you're still not confident, tell the user what you found, what's ambiguous, and what the safest next step is (often: a read-only test, a snapshot first, or asking them to confirm an assumption). Do not guess on destructive operations.
How this composes with safety
Research establishes what the correct action is; the `homelab-safety` skill governs how you get approval to perform it. Use both: research to be right, safety to be safe.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugins/homelab-core/skills/homelab-research/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/redth/homelab-assist/homelab-research)