Loading ToolBench

Reading the current corpus…

hacknotts-cli MCP server · F · ToolBench

← Browse

hacknotts-cli

github.com/Pleasurecruise/hacknotts-cli

Interactive AI-powered CLI for HackNotts 2025 - Chat with multiple AI providers through an elegant terminal interface

2025-06-18+STDIO7 tools

25/100

Definition Quality14

Protocol Readiness32

Supportability39

Spec posture

Inferred effective spec 2025-06-18+·current baseline 2026-07-28

Static source inference · medium confidence · detected: Logging

Deprecated protocol patterns detected

Summary

Critical failures across all dimensions. Tool definitions are inferred from plugin file paths rather than explicitly visible in source code. No input schemas, parameter descriptions, or output schemas are visible in the provided code. Descriptions are generic and under 50 characters. Duplicate tool names (web_search appears twice, google_search appears twice) indicate poor composition and naming clarity. The code_execution tool with WRITE risk has no visible safety constraints, permission gates, or error handling. No evidence of idempotency, pagination, or structured output patterns.

Tools (7)

code_executionwriteauth22/100

Code execution tool from googleToolsPlugin

google_searchread onlyauthsource verified18/100

Google search tool provided by Google

google_searchread onlyauthsource verified18/100

Google search tool from googleToolsPlugin

url_contextread onlyauthsource verified20/100

URL context tool from googleToolsPlugin

web_searchread onlyauth18/100

Web search tool provided by Anthropic

web_searchread onlyauth18/100

Web search tool provided by OpenAI

web_search_previewread onlyauth17/100

Web search preview tool provided by OpenAI

Issues (5)

CRITICALtool ↗

No input schemas visible for any tool. Cannot verify parameter types, constraints, or descriptions.

web_searchweb_search_previewgoogle_searchurl_contextcode_execution

CRITICALtool ↗

Duplicate tool names (web_search appears twice, google_search appears twice) violate composition principle. LLMs cannot disambiguate between identically-named tools. Each tool must have a unique, action-specific name.

web_searchgoogle_search

CRITICALconfirmation-request ↗

code_execution tool (WRITE risk) has no visible permission gates, dry-run support, or confirmation mechanism. Destructive tools must support confirmation_request pattern to prevent catastrophic agent errors.

code_execution

HIGHtool-description ↗

Tool descriptions are generic and under 50 characters (e.g., 'Web search tool provided by OpenAI'). Descriptions must be 10-1024 chars and explain WHAT, WHEN, and any prerequisites. Current descriptions do not guide LLM selection.

Recommendations

Provide explicit tool registration code showing input schemas, parameter types, descriptions, and output schemas. Use JSON Schema with full type definitions and constraints.
Rename duplicate tools to unique, action-specific names. E.g., 'web_search_openai', 'web_search_anthropic', 'google_search_primary', 'google_search_fallback'.
Expand tool descriptions to 50-200 characters. Include WHEN to use each tool and how it differs from similar tools. E.g., 'Search the web using OpenAI's search API. Use this for real-time information on current events, news, and recent data not in training data.'
Add permission gates and dry-run support to code_execution. Implement confirmation_request pattern: return a confirmation token, require explicit approval before executing arbitrary code.
Document output schemas for all tools. Specify which fields are returned, their types, and which fields downstream tools need (e.g., if url_context returns a URL, code_execution must accept that URL as input).

Spec posture evidence

Inferred effective spec: 2025-06-18+.

Relies on Logging (deprecated) - log to stderr or use OpenTelemetry

Score history

Overall score trend

First recorded score · v2 rubric

25/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	25	2025-06-18+	v2