agentleFS
Sign inStart free
Ming00777Ming00777/ai-workflow-advisor/skills/tool-recommendation/SKILL.md

tool-recommendation

Skill1 starsChanged 58 days agoMITView on GitHub

Recommend a small, explainable set of China-focused CLI tools, MCP servers, and Agent products from the local catalog after understanding the user's work goals and authorized workspace audit.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 1 is distinctive, and 4 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.

  • 4

    Very specific

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

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

    Use the catalog under catalog/ as the source of truth.

    line 8
  2. Distinctive:

    Ask what the user is trying to accomplish and what data or collaboration tools they already use.

    line 12
  3. Very specific:

    Use an authorized workspace-audit result if available.

    line 13
  4. Very specific:

    Never invent an official download URL, CLI command, MCP endpoint, current version, or product integration.

    line 17
  5. Very specific:

    Do not auto-install or mutate any system state.

    line 35

≈ “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

Tool Recommendation

Use the catalog under catalog/ as the source of truth. Read only the relevant catalog file for the requested tool type.

Workflow

  1. Ask what the user is trying to accomplish and what data or collaboration tools they already use.
  2. Use an authorized workspace-audit result if available. Never infer a tool choice from file names alone, and never treat a missing manifest or binary as proof that a user does not need that capability.
  3. Match the task to a maximum of five candidates. Rank by task fit, current gap, installation feasibility, source quality, permission cost, and maintenance status.
  4. Explain each recommendation: problem solved, why it matches the user's stated goal or a clearly labeled hypothesis, prerequisites, permission scope, official or community status, and a safe verification step.
  5. Include one or two lower-risk alternatives and say why they were not ranked first.
  6. Never invent an official download URL, CLI command, MCP endpoint, current version, or product integration. If the catalog says needs_verification, recommend checking official documentation instead.

Recommendation classes

  • official: maintained by the vendor or documented on its official site;
  • community: useful third-party project, clearly labeled and subject to source review;
  • product_only: an Agent product with no verified public CLI/MCP installation path in this catalog;
  • needs_verification: candidate found during research but not safe to present as installable.

Output format

  1. Current user goal
  2. Observed facts versus assumptions
  3. Ranked recommendations
  4. Installation order
  5. Permissions and privacy notes
  6. What to verify before installation

Do not auto-install or mutate any system state. Installation belongs to setup-guide and still requires explicit authorization immediately before execution.

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.

Check your setup, free