Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Mcp Tooling Fit

ASecurity

Select the smallest useful MCP/tool set for the current Agentit task, based on repository facts, explicit capability needs, least privilege, and current availability.

3 stars
0 votes
0 copies
0 views
Added 9/19/2026
ai-agentsbashfrontendbackenddocumentation

Works with

mcp

Security Analysis

A100/100

Scanned 9/19/2026

Install to Claude Code

$npx -y skills add marcmarti9/agentit --skill mcp-tooling-fit --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Mcp Tooling Fit?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Mcp Tooling Fit
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/marcmarti9-mcp-tooling-fit/badge)](https://www.skillsdirectory.com/skills/marcmarti9-mcp-tooling-fit)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
name: mcp-tooling-fit
description: Select the smallest useful MCP/tool set for the current Agentit task, based on repository facts, explicit capability needs, least privilege, and current availability.
---

# MCP Tooling Fit

Choose tools for the reviewed task, not because they are popular or globally installed. The primary AI owns the semantic choice; deterministic code may resolve an explicit server/stack ID and enforce risk gates.

## When to use

Load this skill JIT when the task materially depends on MCP/tool selection, when the current tool inventory is noisy, or when an important capability appears missing.

Do not load it for ordinary work whose existing tools are already sufficient.

## Protocol

### 1. Inspect current state

Agent-facing helpers include:

```bash
agentit mcp status --project .
agentit mcp available --project .
agentit mcp active --project .
```

The agentit-manager MCP can expose the equivalent inventory/enable/disable operations.

Record only what matters for the current task:

- explicit current availability;
- required permissions/capabilities;
- write vs read-only behavior;
- secrets/env requirements;
- risk classification.

Catalog presence is not runtime availability.

### 2. Infer needs from task and repository facts

Inspect manifests, frameworks, DB/config, browser/UI surfaces, design sources, CI/deployment and external integrations. Do not ask the user for facts already visible in the project.

Example curated stacks may include:

| Need | Candidate stack |
|---|---|
| repository/docs/browser baseline | `developer_core` |
| frontend/browser verification | `frontend` |
| design inspection | `design_studio` |
| mobile / Expo / RN UI | `mobile_design` (appllama + context7; paid credits) |
| backend/data work | `backend_data` |
| current-source research | `research` |
| product operations | `product_ops` |

These names are discovery conveniences, not semantic routing rules. The primary AI decides whether a stack actually fits.

### 3. Prefer the minimum useful set

Keep or enable only tools that earn their permission/context/coordination cost. Disable unrelated always-on noise when doing so is safe and useful.

Plan mutations before applying them. Higher-risk write-capable tools require the applicable human/review gate.

### 4. Discover gaps when necessary

If Agentit's catalog does not cover a required capability:

1. prefer official/maintained providers;
2. check source, maintenance, scopes and secret handling;
3. compare a small number of realistic candidates;
4. do not silently install or authorize external software.

Current-source discovery belongs in the task's reference/tool plan, not in a permanent global preload.

### 5. Persist only operationally useful state

When a substantial task needs continuity, record the selected MCP/tool set and pending approvals in local `.agentit/STATE.md`. Do not publish transient tool enablement decisions as tracked project documentation unless they are a durable architectural/operational contract.

## Safety

- least privilege by default;
- secrets through environment/provider secret stores, never repository files;
- verify current auth/availability before depending on a mutable external service;
- RISK_3/RISK_4 or destructive external actions require the applicable human and independent-review gate;
- disabling/re-enabling tools is allowed when reversible and authorized.

## Anti-patterns

- enabling every MCP "just in case";
- treating a stack name as a natural-language router;
- assuming catalog presence means authenticated availability;
- silently installing third-party servers;
- asking for stack facts that the repository exposes;
- writing secrets or mutable tool state into public docs;
- using powerwords to activate tooling.

## Verification

- [ ] current inventory was inspected when material;
- [ ] selected tools map to explicit task capabilities;
- [ ] unnecessary permission/context noise was avoided;
- [ ] external installs/actions respected risk and consent gates;
- [ ] substantial continuity notes, if needed, remain in `.agentit/STATE.md` by default.

Attribution

marcmarti9marcmarti9
View sourceMore from marcmarti9 →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.

1023331 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

686011 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3331 votes

catchup

Recovers prior coding-agent session context by running `catchup <agent> --since-compact`, which extracts a clean summary of a previous Codex, Claude Code, Antigravity, OpenCode, or Pi Agent session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", or asks to recover/summarize a previous session before continuing. Do NOT use for the current conversation, git history, or any non-agent log.

611 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →