Unverified50% confidenceBenchmarkExact time
The two Skills in Origin Apps occupy approximately 0 tokens when resident, and each invocation costs fewer than 20 tokens, whereas a self-written system prompt typically requires several thousand tokens.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/17/2026
First Seen
Valid until: 10/31/2026
Sources
Dissecting Cursor's 10K-Star Plugin Repository: Install into Claude Code with One Command
bilibili1024工程笔记10/15/2026
Related Entities
Related Claims
UnverifiedToken generation is billed by consumption; if an application requests a 4,000-token completion and the user stops after the 100th token, failure to propagate the cancellation signal means paying for the remaining 3,900 unrendered tokens.64% similarUnverifiedIn the effort setting, a mid-low-tier code review writes 470 tokens while a high-tier one writes 1,200 tokens; the low tier is 65% cheaper and 5 seconds faster, but setting effort for sub-agents yields no benefit and instead doubles the cost.64% similarUnverifiedThe prefill throughput of this setup increases with prompt length, climbing from approximately 290 tokens/s for short prompts to approximately 1,300 tokens/s, with an average prefill throughput of approximately 900 tokens/s.63% similarUnverifiedOverly long system prompts bring hidden overhead: for example, when the system prompt has 500 tokens while the actual input has only 50 tokens, it means 91% of the cost goes to boilerplate text; auditing and trimming the system prompt by 40% can reduce costs.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1056591API
curl https://kongchang.com/api/v1/knowledge/claims/1056591MCP
get_claim(id=1056591)