Public assessment

lulzasaur9192/marketplace-search-mcp (@lulzasaur9192/marketplace-search-mcp)

lulzasaur9192-marketplace-search-mcp · v2.0.0 · scanned Aug 3, 2026

What changed in the harness

Selection accuracy 100→98, token cost up 7%, unconfirmed writes 100%→100%.

Category breakdown

Where the score comes from.

Earned points across the four signals Gradable measures. Safety and Legibility are scored out of 30; Economics and Discoverability are scored out of 20.

01Safety

30.0 / 30

30.0 out of 30

02Legibility

24.8 / 30

24.8 out of 30

03Economics

17.7 / 20

17.7 out of 20

04Discoverability

15.4 / 20

15.4 out of 20

Highest-impact fix

Estimated gain +1 point

Rewrite 12 underspecified tool descriptions

State each tool's behavior, explain every parameter, and declare return semantics, prioritizing tools exercised by failed selection tasks.

Description evidence

Defects and rewrites.

12 defects found across the exposed tool descriptions. Suggested rewrites make purpose, inputs, boundaries, and returns easier for an agent to understand.

Tool	Defect types	Suggested rewrite
`verify_nurse_license`	params_unexplained	Verify a nurse's license across US states (FL, NY) from the official state nursing boards (FL: DOH MQA, NY: NYSED). Search by licenseNumber, by last name (optionally refined with firstName for a more precise person search), or by licenseType (RN, LPN, NP, APRN, CNA). Returns license status, expiration, qualifications, and enforcement actions, up to the given limit.
`search_nyc_violations`	no_return_description	Search NYC Department of Buildings violation records by street address (house_number + street), BIN, or tax block+lot (block + lot), optionally restricted to a borough. Covers DOB Violations, Safety Violations, and ECB Violations. Returns matching violation records with violation type, status, dates, and penalty amounts; set summary=true to return aggregate stats instead of individual violations, and limit (ignored when summary=true) to cap results.
`childcare_states`	no_return_description	Get childcare cost averages for all US states from the DOL National Database of Childcare Prices. Returns each state's average weekly and annual prices for center-based and family childcare by age group, for the given year (defaults to latest available).
`childcare_compare`	no_return_description	Compare childcare costs side-by-side across up to 10 ZIP codes or locations. Returns a per-location breakdown of average weekly and annual prices for center-based and family childcare by age group, for the given year (defaults to latest).
`search_estate_sales`	no_return_description	Search for estate sales by ZIP code or city/state on EstateSales.net. Returns upcoming and active estate sales, each with dates, address, description, photos, and a link to the full listing, up to the given limit.
`estate_sale_details`	no_return_description	Get full details of a specific estate sale from EstateSales.net, identified by its sale_id or url. Returns the sale's photos, listed items, dates, directions, and sale company info.
`search_storage`	params_unexplained no_return_description	Search self-storage facilities by city/state or ZIP code on SelfStorage.com. Returns matching facilities with unit sizes, pricing, and availability. Use page (1-based result page, defaults to 1) together with limit to paginate through the results.
`storage_facility_details`	no_return_description	Get full details for a specific self-storage facility from SelfStorage.com, identified by its facility URL. Returns the facility's unit sizes and prices, amenities, hours, and reviews.
`search_gsa_auctions`	no_return_description	Search US government surplus auctions from GSA (General Services Administration) for vehicles, electronics, equipment, and more at below-market prices. Returns matching auction lots with descriptions, current bid, status, and location. Filter by state, agency, status (active or preview), and min_bid/max_bid; paginate with offset and limit (max 50).
`gsa_auction_details`	no_return_description	Get full details of a specific GSA surplus auction lot, identified by sale_number and lot_number. Returns the lot's description, bid info, location, inspection dates, and sale terms.
`gsa_auction_stats`	no_return_description	Get aggregated statistics for GSA surplus auctions. Returns overall counts and breakdowns by state, agency, and auction status, along with total bid amounts and averages.
`skillguard_verify`	no_return_description	Verify the safety of an AI agent skill/tool before execution by analyzing its name, description, permissions, and commands. Returns a safety classification of SAFE, CAUTION, or DANGER along with the findings that justify it. Use this to audit MCP tools, OpenAI functions, or any agent capability.

Selection evidence

Confusable tool pairs.

6 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.

Tool A	Tool B	Confidence	Why they collide
`search_fcc`	`fcc_details`	high	Both accept an FCC ID and both descriptions promise 'certification details'; a task like 'get the details for FCC ID BCG-E3994A' can plausibly route to either, since search_fcc already returns equipment class and grant dates while fcc_details only adds full documents.
`search_houzz`	`search_thumbtack`	high	Both search local service professionals/contractors with ratings and reviews; a task like 'find contractors in San Francisco' fits both (Houzz explicitly covers contractors/interior designers, Thumbtack covers 1000+ service categories), differing mainly in input style (URL vs query).
`verify_contractor_license`	`verify_nurse_license`	medium	Both verify a person's professional license by name/number in overlapping states (FL, NY) and return status/expiration; a task like 'verify John Smith's license in Florida' is ambiguous when the profession is not stated.
`childcare_cost`	`childcare_compare`	medium	Both return childcare prices by ZIP; compare accepts one or more ZIPs, so a single-location request like 'what does childcare cost in 90210' could be routed to either, and only multiple-location phrasing clearly favors compare.
`childcare_cost`	`childcare_states`	medium	Both describe childcare cost lookups keyed to location; 'average childcare cost by state' is ambiguous between a per-state lookup (cost with a state param) and the cross-state all-states comparison tool.
`gsa_auction_details`	`gsa_auction_stats`	medium	Both concern GSA surplus auctions and 'details' can describe either per-lot info or aggregate stats; a task like 'give me details/overview of GSA auctions' without a specific sale/lot number could route to either.

Compare the field

One score is useful.
The evidence makes it actionable.

Back to the leaderboard