CtrlK
BlogDocsLog inGet started
Tessl Logo

mcp-builder

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

78

1.60x
Quality

71%

Does it follow best practices?

Impact

88%

1.60x

Average score across 3 eval scenarios

SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/data/0b1c2d3e-4f5a-6b7c-8d9e-0f1a2b3c4d5e/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

56%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, phase-by-phase process guide with clear workflow sequencing and validation, undermined by an absent reference bundle: all `./reference/*.md` links are broken and no executable implementation code is supplied in the body, plus some redundant restatement of load instructions.

Suggestions

Ship the referenced bundle (mcp_best_practices.md, python_mcp_server.md, node_mcp_server.md, evaluation.md) under ./reference/ — or replace the dead links with inline minimal examples — so the deferred detail is actually reachable.

Remove the duplicated 'Reference Files' section at the end and keep load instructions only at their first point of use in Phase 1/2, since the repetition adds tokens without new information.

Add at least one small copy-paste-ready tool example inline (Pydantic input model + @mcp.tool registration, or a Zod schema + server.registerTool) so the body is actionable even before the reference files are loaded.

DimensionReasoningScore

Conciseness

Mostly bullet-driven and reasonably lean, but the final 'Reference Files' section restates load instructions already given in Phases 1.4/2.4, the Overview restates the description, and Phase 1.1's design principles pad general AI-tool advice Claude already knows — enough redundancy to keep it at the midpoint rather than 4.

3 / 5

Actionability

Concrete commands appear (WebFetch URLs, `python -m py_compile`, `npm run build`, `timeout 5s python server.py`, tool-annotation fields), but the core implementation guidance is deferred to `./reference/*.md` files that do not exist in the bundle, leaving key executable detail missing.

3 / 5

Workflow Clarity

Four phases are clearly sequenced and Phase 3 supplies validation checkpoints (syntax check, build, quality checklist, test-safety warnings) with implicit fix-and-retry guidance; minor validation gaps (no explicit re-run loop) keep it just below 5.

4 / 5

Progressive Disclosure

References are well-signaled and one level deep with phase-organized navigation, but every `./reference/*.md` path points to a directory that does not exist in the bundle, so the disclosure promises detail that is not actually available — a material gap that pulls it down from 4.

3 / 5

Total

13

/

20

Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description that clearly states both capability and trigger conditions with concrete, distinguishable language. The only weakness is mild generic padding ('high-quality', 'well-designed') and a few missing trigger synonyms.

DimensionReasoningScore

Specificity

Names the domain (MCP servers) and several concrete actions — 'creating high-quality MCP servers', 'interact with external services through well-designed tools', 'integrate external APIs or services' — with only minor generic padding like 'high-quality'/'well-designed' keeping it below a 5.

4 / 5

Completeness

Explicitly answers both 'what' (guide for creating MCP servers that enable LLMs to interact with external services through well-designed tools) and 'when' ('Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK)') with concrete trigger phrases.

5 / 5

Trigger Term Quality

Includes natural phrases users would say ('building MCP servers', 'integrate external APIs or services') plus synonyms/frameworks (Python/FastMCP, Node/TypeScript/MCP SDK), but a few common variations are absent, so it sits just below comprehensive.

4 / 5

Distinctiveness Conflict Risk

Occupies a clear niche (MCP server construction) with distinct triggers (MCP servers, external APIs/services, FastMCP, MCP SDK) and minimal realistic overlap with other skills.

5 / 5

Total

18

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 14 missing

Warning

Total

15

/

16

Passed

Repository
majiayu000/claude-skill-registry
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.