csh/playbook
1
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
16 Commits 3 Branches 0 Tags
9d2b696f58dbed31a514c748aec4a480826d288c
Commit Graph
7 Commits
Author SHA1 Message Date
cshandClaude Opus 4.8 efd4b8e421 📝 docs(tsl): align README with syntax skill, restructure roadmap as part-3 
- rewrite part-1 to reflect tsl-syntax-reference skill (lookup retrieval, layered table, dual-skill routing)
- fix fabricated section ID in README and SKILL.md examples (verified via --section)
- trim block-identity table, api-reference counts, and stale compare-phrasing
- promote roadmap to standalone part-3 with four directions incl. Agent Loop (planned)
- add deployment section (per-agent skills dirs) and AGENT_PRIMER.md primer

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
 2026-07-14 09:09:58 +08:00
csh ce6698354a 📝 docs(tsl): update README subtitle 2026-07-08 15:44:37 +08:00
cshandClaude Fable 5 23779c95ab 📝 docs(tsl): restructure part-2 benchmark section per review 
- 五个考察点编号①-⑤并贯穿各轮回放表与横向对比，补「语法证据」维度
- 三轮小节统一模板：环境行 / 回放表 / 产出(存档链接) / 亮点 / 对照差异 / 成本口径
- 提示词差异与第二轮已读过第一轮示例改为就地披露
- 新增「三轮小结」回链后续升级方向；节名「实际场景」改「多环境实测」

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
 2026-07-08 13:33:45 +08:00
cshandClaude Fable 5 296ea0fc40 📝 docs(tsl): apply README review fixes 
- builtin 分类 12 类 -> 10 类，注明 finance/market 空壳类目并纳入分类治理方向
- 首跳 mermaid 拆分「写代码 / 查语法」两条边，与 docs/tsl/index.md 实际路由一致
- TOC 补齐第二部分各小节锚点；Token 高效改标为设计目标（估算）
- 目录树补 scripts/ 存档行；对照说明记录第一轮产出加 _claude 后缀的改名
- 横向对比「5/5」改为「skill 逐一查证 5/5」统一口径

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
 2026-07-08 12:49:31 +08:00
cshandClaude Opus 4.8 061e67eb06 📝 docs(tsl): archive three benchmark scripts referenced by README 
- add scripts/main_board_top50_zf_{claude,codex,qwen3.6-27b}.tsl as produced
- sync round-1 code block and archived filename in README

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
 2026-07-08 12:26:20 +08:00
cshandClaude Opus 4.8 76a6523ac6 📝 docs(tsl): record three-agent benchmark in README 
- rewrite Part 2 as real tests: Claude Fable 5 / Codex GPT-5.5 / Qwen3.6-27B
- add cross-env comparison table with per-round token caliber caveats
- backfill roadmap evidence: lookup.py encoding fix, equity/misc.md 9252-line pile-up

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
 2026-07-08 12:21:22 +08:00
cshandClaude Fable 5 e3fd9625dc 📝 docs(tsl): add TSL playbook README 
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
 2026-07-08 11:34:06 +08:00