agentleFS
darellchua2darellchua2/civiltekk-opencode-claude-skills/skills/install-assistant-skill/SKILL.md

install-assistant-skill

Skill7 starsChanged 56 days agoApache-2.0View on GitHub

Guided install assistant — one entry point for MCP servers (inventory, precheck, --enable-pack or per-project enable, verify) and the civiltekk catalog (find, install with mandatory dry-run + consent, update, remove). Routes each install task to the right specialist route. Never auto-installs.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Verify after applying (opencode mcp list / installer listing).

    line 42
  2. Very specific:

    Never hand-edit installed copies or global config — fix at the source repo, then update.

    line 44
  3. Distinctive:

    Never advise enabling an MCP whose endpoints are not verifiable from the vendor's own docs (the Autodesk precedent, #554).

    line 59

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Install Assistant Router

One entry point for "install/enable X" tasks. Identify the surface, run only the matched route's file, and follow it. Consent is the spine: dry-run → show the resolved set → user consent → apply → verify. Never auto-install.

Provenance: consolidated from mcp-install-assistant-skill and civiltekk-install-assistant (#691); the catalog route's find/install/maintain values files predate it (#657 lineage in route files).

Routes (load exactly one)

TaskRouteReadSkip-when
Enable/diagnose MCP servers (packs, keys, browsers)mcpreferences/routes/mcp.mdcatalog skills → catalog
Find/install/update/remove catalog skills + subagentscatalogreferences/routes/catalog.mdMCP servers → mcp

Route files open with their own Load when / Skip when lines; the catalog route's detail files (catalog/find|install|maintain.md) load per its own table — never all three.

Consent spine (both routes)

  • Dry-run before every write; show the resolved set (entry + transitive requirements) before the real add/enable.
  • Global (user-scope) changes require consent; project scope is the default.
  • --no-deps only on an explicit user ask — never proactively.
  • Verify after applying (opencode mcp list / installer listing); report exact errors verbatim — no blind retries, no invented flags.
  • Never hand-edit installed copies or global config — fix at the source repo, then update; the deploy script owns global config.

Capability bindings

Interactive selection prompts: OpenCode — question tool (one question, 2–4 options, multiple allowed). Claude Code — AskUserQuestion. Other/none — number the options in a plain reply and wait for the pick. Requires bash (git-bash/WSL on Windows) for enable/verify steps; the inventory is plain Node.

Hard rules

  • Atlassian ships disabled globally and must not be force-enabled globally (project opt-in only).
  • Never advise enabling an MCP whose endpoints are not verifiable from the vendor's own docs (the Autodesk precedent, #554).
  • Full-stack setup requests are a pointer, not an install: route to the deploy surface (civiltekk-setup); only the catalog route installs skills.
  • Each route file owns its detail; this router never restates it.

Gates

  • Zero-preload: only the matched route file (+ the catalog route's named detail file) enters context.
  • Report shape complete: what changed, where (scope), verification output, skipped items and why.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.