CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

setup-tooluniverse

Install and configure ToolUniverse for any use case — MCP server (chat-based), CLI (command line with 14 subcommands), or Python SDK (Coding API with 3 calling patterns). Covers uv/uvx setup, MCP configuration for 12+ AI clients (Cursor, Claude Desktop, Windsurf, VS Code, Codex, Gemini CLI, Trae, Cline, etc.), full CLI reference (tu list/grep/info/find/run/test/status/build/remote/doctor/serve/connect/connections/disconnect), Coding API quickstart, agentic tools, code executor, API key walkthrough, skill installation, and upgrading. Use when user asks how to set up ToolUniverse, which access mode to use (MCP vs CLI vs SDK), configuring MCP servers, using the CLI, troubleshooting installation, upgrading, or mentions installing ToolUniverse or setting up scientific tools. Also triggers for "how do I use ToolUniverse", "what's the best way to access tools", "command line", "tu command", "coding API", "tu build".

Quality

91%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

85%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 strong, highly actionable setup guide: nearly everything is copy-paste executable, the multi-mode workflow is clearly sequenced with validation checkpoints and error-recovery paths, and the prose is mostly tight. The main defect is progressive disclosure — a broken reference to a missing API_KEYS_REFERENCE.md and four bundled scripts that are never linked from the body — plus a few novice-oriented explanation blocks that could be trimmed.

Suggestions

Fix the broken reference at 'See [API_KEYS_REFERENCE.md](API_KEYS_REFERENCE.md) for the complete list with all tiers' — either add the file to the bundle or inline the remaining key tiers, since following the link dead-ends.

Reference the four scripts in scripts/ (check_prerequisites.py, diagnose_setup.py, list_tool_categories.py, verify_installation.py) from the relevant sections (e.g., Step 4 Test Together or Common Issues), or remove them from the bundle so no content is orphaned.

Trim the blocks that explain concepts Claude already knows — the terminal how-to ('Mac: Cmd+Space → Terminal'), 'Config files are plain text that store settings', and the 'safe, standard command' reassurance — keeping at most a one-line pointer for novice users.

DimensionReasoningScore

Conciseness

The bulk is token-efficient — dense command tables (CLI subcommands, config locations, key tiers, common issues) and copy-paste blocks with no filler. A few localized blocks over-explain basics Claude already knows: "Terminal help (if needed): Mac: Cmd+Space → 'Terminal' → Enter", "Config files are plain text that store settings — like a preference list for the app", and "(This is a safe, standard command that downloads and installs `uv`, a small package manager...)". These are conditionalized for novice users, keeping this at anchor 4 (efficient with minor trimmable over-explanation) rather than 5; it is above anchor 3 because the unnecessary explanation is confined to a handful of short conditional asides rather than pervasive padding.

4 / 5

Actionability

Guidance is fully executable and copy-paste ready: "curl -LsSf https://astral.sh/uv/install.sh | sh", "uvx --from tooluniverse tu status", the complete mcpServers JSON block, all three runnable Python calling patterns, per-client config file paths, API key registration URLs, test commands ("uvx tooluniverse tu status && tu find 'protein'..."), a validation one-liner ("python3 -m json.tool < <config>"), a demo-query table, and a prompt cheat sheet. This matches the anchor for copy-paste ready commands with specific examples covering the common cases; no pseudocode or missing key details were found.

5 / 5

Workflow Clarity

The flow is a clearly sequenced Step 1–5 process with "Continue to Step 3" signposts and explicit validation checkpoints: "Verify: `uv --version`", "Then validate before restarting the app: python3 -m json.tool", and Step 4 "Test Together" with concrete test calls. Error recovery is well handled: "If issues: Most common: app not restarted, `uv` not in PATH (reopen terminal), JSON syntax error in config", a Common Issues table, and the "Health check: `tooluniverse-doctor`" feedback loop. This matches the anchor for clear sequence with explicit validation steps and feedback loops for error recovery.

5 / 5

Progressive Disclosure

One reference is well handled — "See [references/mcp-configs.md](references/mcp-configs.md)" points to a real one-level-deep file that appropriately offloads special config formats. But the body cites "See [API_KEYS_REFERENCE.md](API_KEYS_REFERENCE.md) for the complete list with all tiers" and that file does not exist in the bundle (broken navigation), and the four scripts in scripts/ (check_prerequisites.py, diagnose_setup.py, list_tool_categories.py, verify_installation.py) are never referenced anywhere in the body, leaving an entire bundle directory undiscoverable. This matches anchor 3 (some structure, but organization/navigation gaps are more than minor); it is not anchor 4 because a broken reference and an orphaned bundle directory are real navigation failures, and not anchor 2 because the body itself is well-sectioned and one reference is properly split out.

3 / 5

Total

17

/

20

Passed

Description

96%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.

An exemplary description: it explicitly covers what the skill does across all three access modes with named subcommands and clients, and gives both formal 'Use when...' conditions and quoted natural trigger phrases. The only weakness is a couple of unqualified generic triggers ("command line", "configuring MCP servers") that create minor overlap risk with other setup-related skills.

DimensionReasoningScore

Specificity

The description lists multiple concrete actions with comprehensive coverage: "Install and configure ToolUniverse for any use case — MCP server (chat-based), CLI (command line with 14 subcommands), or Python SDK (Coding API with 3 calling patterns)", plus "uv/uvx setup, MCP configuration for 12+ AI clients", "full CLI reference (tu list/grep/info/find/run/test/status/build/remote/doctor/serve/connect/connections/disconnect)", "API key walkthrough, skill installation, and upgrading". This matches the anchor for multiple specific concrete actions with comprehensive coverage; it is not score 4 because no meaningful capability area is missing — all three access modes, the complete command list, keys, skills, and upgrades are each explicitly named.

5 / 5

Completeness

Both questions are explicitly answered. "What": "Covers uv/uvx setup, MCP configuration for 12+ AI clients..., full CLI reference..., Coding API quickstart, agentic tools, code executor, API key walkthrough, skill installation, and upgrading." "When": an explicit "Use when user asks how to set up ToolUniverse... or mentions installing ToolUniverse or setting up scientific tools" clause followed by "Also triggers for..." with concrete phrases. This matches the anchor for clearly and explicitly answering both what and when with concrete trigger phrases; the 'when' clause is present and specific, so the completeness cap of 3 for a missing 'Use when...' clause does not apply.

5 / 5

Trigger Term Quality

Trigger coverage is comprehensive and natural, mixing formal conditions ("Use when user asks how to set up ToolUniverse, which access mode to use (MCP vs CLI vs SDK), configuring MCP servers, using the CLI, troubleshooting installation, upgrading") with quoted colloquial phrasings users would actually say ("how do I use ToolUniverse", "what's the best way to access tools", "command line", "tu command", "coding API", "tu build"). This matches the top anchor for comprehensive natural-term coverage including synonyms; it is not score 4 because both formal trigger clauses and natural user phrasings are present with few gaps.

5 / 5

Distinctiveness Conflict Risk

Most triggers are anchored to the named product ("installing ToolUniverse", "tu command", "tu build"), giving a clear niche with minimal conflict risk. However, generic phrases like "command line" and "configuring MCP servers" could overlap with other MCP-setup or CLI-related skills, which matches the anchor for mostly distinct with minor overlap risk. It is not score 5 because those generic trigger terms are not product-qualified; not score 3 because the skill is overwhelmingly anchored to a specific named product.

4 / 5

Total

19

/

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.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 1 missing

Warning

Total

15

/

16

Passed

Repository

mims-harvard/ToolUniverse

Path

plugin/skills/setup-tooluniverse/SKILL.md

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.