Unverified50% confidenceFactExact time
The agent used 19 headless runs over two days to determine that Claude Code's MEMORY.md only loads the first 200 lines or 25,000 bytes (whichever is smaller), truncating at line boundaries and appending a warning
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/11/2026
First Seen
Valid until: 1/9/2027
Sources
Related Entities
Related Claims
UnverifiedIn preliminary multi-agent experiments given a full 8 hours, the programming benchmark peaked at only 30.04 (versus 20.39 for single-agent), and the science-category agent benchmark reached only 31.2, lagging behind a competitor's 51.8.67% similarUnverifiedAfter bypass mode was enabled, Devin worked independently for approximately 10–15 minutes, generating or modifying 21 files and 277 lines of code, and completed content population and testing in the local database.65% similarUnverifiedPilot experiments had three coding agents and eight models each run eight 20-minute single-pass tasks, with label-aware F1 showing extremely high variance ranging from 0.143 to 0.892.64% similarUnverified在实现车辆驾驶模拟器测试中，千问3.8 27B开启深度思考模式后思考了将近29分钟，生成约600多行代码64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1000308API
curl https://kongchang.com/api/v1/knowledge/claims/1000308MCP
get_claim(id=1000308)