Skip to content

docs: 隱私措辭誠實化 + Step 4 反饋落地管道 (closes #42, closes #43)#70

Open
atomchung wants to merge 1 commit into
mainfrom
claude/issue-42-43-privacy-feedback
Open

docs: 隱私措辭誠實化 + Step 4 反饋落地管道 (closes #42, closes #43)#70
atomchung wants to merge 1 commit into
mainfrom
claude/issue-42-43-privacy-feedback

Conversation

@atomchung

Copy link
Copy Markdown
Owner

摘要

兩個 P0 合一 PR(都動 README 隱私節同一區域,拆開必衝突):

#43 隱私措辭(#48 之後的殘餘,依 issue comment 的增量三點)

#42 反饋迴路落地

  • .github/ISSUE_TEMPLATE/card-feedback.yml:兩格(有戳中嗎?/哪裡不對?)+ 真實 CSV vs mock 下拉(Stage 0 關鍵訊號);表單頂部明示別貼交易明細
  • SKILL.md Step 4:收完反饋給一個可點的表單連結,自願、只提一次、不推銷
  • README 隱私節同步「唯一自願回收的反饋怎麼給」
  • label card-feedback 已建好(gh 實測存在)

驗收(對照 #42)

  • Step 4 收尾真的給出可點的回饋入口(URL 寫死在 SKILL.md,merge 後 template 生效)
  • template 不誘導貼交易內容(兩層警語:SKILL 話術 + 表單 markdown)
  • YAML 過 safe_load 驗證;python3 tests/run_all.py 3 套全綠

Closes #42, closes #43

🤖 Generated with Claude Code

, closes #43)

Privacy (#43, residual after #48):
- README privacy section rewritten architecture-neutral: data-flow
  self-evidence line (local CSV -> your own model session -> local
  ~/.trade-coach/, no fourth stop); inference goes through the user's
  own model API, subject to their account's data policy
- Drop ambiguous claim about memory; keep the true claims (no
  third-party backend, author never sees a row)

Feedback channel (#42):
- Add .github/ISSUE_TEMPLATE/card-feedback.yml: two text fields
  (hit? / what's off?) + real-CSV-vs-mock dropdown; template and form
  both warn against pasting trade details; card-feedback label created
- SKILL.md Step 4: after collecting the one-line feedback, show a
  clickable form URL once, voluntary, no second pitch
- README privacy section links the same form as the only voluntary
  collection channel (Stage 0's source of truth)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

1 participant