Loading ToolBench

Reading the current corpus…

fundzwatch-python MCP server · F · ToolBench

fundzwatch-python

Python SDK for FundzWatch.ai - Real-time business event intelligence for AI agents. Provides access to AI-scored leads, business events, market intelligence, and company watchlist management.

2026-07-28+8 tools

40/100

Definition Quality53

Protocol Readiness0

Supportability45

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

This SDK provides 8 tools across CrewAI and LangChain integrations, but exhibits significant quality gaps. Tools 1-4 and 5-8 are duplicates (CrewAI vs LangChain variants). Naming is verb-forward (get_, manage_) but descriptions lack depth, most are 50-100 chars, below the 194-char baseline for A+ tools. Parameter schemas are present with types and defaults, but descriptions are minimal (e.g., 'Comma-separated stages' without explaining what each stage means or when to use them). No output schemas documented. Error handling is absent from tool definitions. The SDK is a Python client library, not an MCP server, no evidence of MCP protocol implementation (no transport, no tool registration via MCP, no _meta handling). This is a fundamental mismatch: the evaluator requested an MCP server assessment, but the artifact is a REST API wrapper.

Tools (8)

fundzwatch_eventsread onlyauth50/100

Get real-time business events: funding, acquisitions, hires, contracts, launches.

fundzwatch_market_pulseread onlyauth50/100

Get real-time market overview: funding totals, acquisitions, exec moves.

fundzwatch_scored_leadsread onlyauth50/100

Get AI-scored sales leads matched to your ICP. Returns companies with buyer intent scores, buying stage, and outreach recommendations.

fundzwatch_watchlistwriteauth50/100

Manage company watchlist: 'list', 'add' domains, or 'remove' domains.

get_business_eventsread onlyauth50/100

Get real-time business events: funding rounds, acquisitions, executive hires, government contracts, and product launches.

get_market_pulseread onlyauthsource verified63/100

Get a real-time market overview: funding totals, acquisition counts, executive moves, and largest rounds for the past week.

get_scored_leadsread onlyauth50/100

Issues (5)

CRITICALmcp-protocol ↗

Not an MCP server, this is a REST API Python SDK. No MCP protocol implementation detected (no transport, no tool registration via MCP spec, no stateless request handling). Cannot be evaluated as an MCP server.

all

HIGHtool ↗

Duplicate tool definitions: tools 1-4 (CrewAI) and 5-8 (LangChain) are functionally identical. This wastes agent reasoning cycles and violates single-responsibility principle. Consolidate into one canonical set.

get_scored_leadsfundzwatch_scored_leadsget_business_eventsfundzwatch_eventsget_market_pulsefundzwatch_market_pulsemanage_watchlistfundzwatch_watchlist

HIGHtool-description ↗

Descriptions are too brief (50-100 chars vs 194-char baseline). E.g., 'Get real-time market overview: funding totals, acquisition counts, executive moves, and largest rounds for the past week.' lacks WHEN to use it, what prerequisites exist, or what the output structure is.

get_market_pulsefundzwatch_market_pulse

Recommendations

Wrap this SDK in an actual MCP server implementation (e.g., using mcp-python or similar). Expose tools via the MCP protocol with proper tool registration, stateless request handling, and _meta fields per the 2026-07-28 spec.
Remove duplicate tool definitions. Keep only one set (e.g., get_scored_leads, get_business_events, get_market_pulse, manage_watchlist) and remove the fundzwatch_* variants. This reduces agent confusion and reasoning overhead.
Expand tool descriptions to 150-250 characters. Include: what the tool does, when to use it (vs similar tools), what it returns, and any prerequisites. Example: 'Retrieve AI-scored sales leads matched to your ICP with buyer intent scores (0-100), buying stage, and outreach recommendations. Use this to identify high-priority prospects. Returns up to 50 leads per call.'
Document output schemas for each tool. Specify field names, types, and structure. Example for get_scored_leads: 'Returns {signals_found: int, signals: [{company: string, score: int, stage: string, pain_points: [string], recommendation: string}], metadata: {total: int, returned: int}}'.
Add error handling guidance to tool descriptions. E.g., 'If no leads match your filters, try lowering min_score or removing industry filters. If rate limit is hit, wait 60 seconds before retrying.'

Score history

Overall score trend

First recorded score · v2 rubric

40/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	40	2026-07-28+	v2