docs(roadmap): add #414 — status git_state:"clean" in non-git dir; indistinguishable from clean repo

[code-yeongyu]

Pinpoint #414

status --output-format json returns git_state:"clean" from a directory that is NOT a git repository — same value as a real clean git repo.

Evidence (non-git dir /tmp/probe-nogit):

{"git_state":"clean","git_branch":null,"project_root":null,"changed_files":0,"staged_files":0,"unstaged_files":0,"untracked_files":0}

Evidence (clean git repo):

{"git_state":"clean","git_branch":"main","project_root":"/path/to/repo","changed_files":0,...}

Automation must heuristically parse git_branch:null to detect non-repo state — the structured git_state field gives no signal.

ROADMAP-only PR. No source/test changes.

[Yeachan-Heo]

Review verdict: REQUEST_CHANGES

I inspected the exact diff against origin/main: this is docs-only (ROADMAP.md | 2 ++) and CI is green, so code/runtime risk is low.

Blocker: the new #414 entry duplicates existing ROADMAP #125. #125 already covers the same core issue: non-git directories report git_state:"clean" with git_branch:null / project_root:null, making status/doctor JSON ambiguous. It also already includes root-cause/fix-shape context around the git workspace summary default/headline behavior.

Requested change: replace #414 with a short fresh-run follow-up appended to #125, or rework the entry only if there is a genuinely new non-duplicative contract gap beyond #125.

Summary:

• Docs-only scope: ✅
• CI green: ✅
• Standalone wording actionable: ✅
• Non-duplicative ROADMAP entry: ❌ duplicates 合影合影，国产即将崛起 #125

—
[repo owner's gaebal-gajae (clawdbot) 🦞]

[code-yeongyu]

Closing — duplicate of ROADMAP #125 (already documents in non-git directories). Thanks to gaebal-gajae for catching this in review.