在 D:\Research\DoneBench 做只读代码/报告侦察。目标:找出已有的 oracle-spec、token-matched、cross-f…

Metadata

  • Stable ID: codex-user-prompt:a14158167ab588a4
  • Source kind: codex-session-user
  • Category: research-workflow
  • Timestamp: 2026-05-09T00:17:37.178Z
  • Semantic hash: a14158167ab588a4079a13a35af4836e540096b3da957e24bb843ab07dcf114e
  • Public handling: selected full-text prompt with secret filtering.

Prompt Text

在 D:\Research\DoneBench 做只读代码/报告侦察。目标:找出已有的 oracle-spec、token-matched、cross-family model slice、pass@k/reliability、near-miss breakdown、paper table refresh 相关代码和配置。请读 configs、donebench/scripts、donebench/agents、paper/tables、reports/full_runs/runs/topconf_deepseek_toolplan_full/paper_tables。不要修改文件。输出:哪些已有可复用,缺什么,建议最小实现路径。

Reuse Notes

  • EXTRACTED: This is a selected Codex prompt or automation prompt from the local Codex corpus.
  • INFERRED: Future agents can reuse its structure, constraints, and acceptance criteria when creating similar Codex workflows.