Unverified50% confidenceFactExact time
After mixing 25 to 55 tools from other servers into the candidate set, not a single request was misdirected to another server's tools, and accuracy remained between 91% and 98%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/15/2026
First Seen
Valid until: 1/13/2027
Sources
Related Entities
Related Claims
UnverifiedTools scoring below 40 were correctly selected 95.5% of the time, while tools scoring 60 to 79 instead reached 100%.67% similarUnverifiedFor comparison, the Program-of-Thought baseline achieves 72.0% overall accuracy and the tool-calling agent achieves 58.7%.66% similarUnverifiedA total of 673 run attempts: 542 completed, 97 stopped due to vendor rate limits (i.e. HTTP 429, 14.4%), 12 sent back for review, 10 requested changes, 8 blocked, 2 crashed, 2 still in progress.65% similarUnverifiedIn the illustrative sample constructed in the article, a certain tool's first-call success rate was 76% on the claude-desktop client and 21% on cursor, with an average of 62% across all clients.64% similarUnverifiedAmong the 1,274 MCP tools, 497 tool descriptions scored below 40, and only 18 reached 80 or above.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1035658API
curl https://kongchang.com/api/v1/knowledge/claims/1035658MCP
get_claim(id=1035658)