Unverified50% confidenceFactExact time
The tester lists three limitations: each skill has only three plain-text tasks written by the testing party; tool calls, media generation and deployment services were not tested; and there is only a single judge model.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/19/2026
First Seen
Valid until: 1/17/2027
Sources
Related Entities
Related Claims
UnverifiedIn the demonstration, the test instruction issued by the operator to the Agent contained three elements: clearly defining the task and providing login credentials, referencing a pre-built skill library, and limiting the scope to non-destructive verification only.73% similarUnverifiedThe evaluation method is three plain-text tasks corresponding to each skill, with each task answered both with and without the skill, each set repeated 3 to 10 times, and each answer scored 5 times.71% similarUnverifiedThe reasoning tier should be treated as a parameter to be tested rather than a default value; reasoning is a switch for handling difficult tasks, not a setting to be left permanently on.70% similarUnverifiedAn effective test case generation prompt contains four core elements: role setting, real context, specific task, and output format.69% similarUnverifiedThe skill file sets hard rules as guardrails: only MCP may be used, no test results may be fabricated, and a bug must never be created when tests pass.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1065274API
curl https://kongchang.com/api/v1/knowledge/claims/1065274MCP
get_claim(id=1065274)