Skip to content

Make /rebase risk-based: auto-proceed on no overlap, menus otherwise - #9

Merged
gering merged 5 commits into
mainfrom
fix/rebase-risk-based-confirm
Jun 15, 2026
Merged

Make /rebase risk-based: auto-proceed on no overlap, menus otherwise#9
gering merged 5 commits into
mainfrom
fix/rebase-risk-based-confirm

Conversation

@gering

@gering gering commented Jun 12, 2026

Copy link
Copy Markdown
Owner

Summary

  • /rebase no longer asks for confirmation when the rebase is risk-free: if the files changed on base and on the branch don't overlap since the merge-base, it rebases + force-pushes (--force-with-lease) immediately — invoking the skill is the authorization
  • When files DO overlap, the decision is presented as a real selection menu (AskUserQuestion) instead of a free-text y/n/d prompt
  • All other decision points in the skill (auto-stash, conflict resolution) also become selection menus

Changes

  • skills/rebase/SKILL.md: step 5 rewritten (overlap check via comm -12 on both git diff --name-only sides; safe path vs decision path), step 6 stash menu (refuses to stash silently in --auto), step 7 conflict menu, notes + description updated
  • README.md: force-push safety section + design principles ("Interactive by default, silent when risk-free")
  • pr-flow 1.2.1 → 1.2.2 (plugin.json + marketplace.json in sync)

Readiness

  • ✅ scripts/check-structure.py: 0 errors (2 pre-existing warnings, none added)
  • ✅ README updated
  • ✅ Version bumped (patch)
  • ➖ Changelog: repo has none yet (tracked task add-changelog)
  • ➖ Tests/lint/build: N/A (markdown-only plugin)

Test plan

  • Run /rebase on a branch with zero file overlap vs base → proceeds without prompt, aborts cleanly on unexpected conflict
  • Run /rebase with overlapping files → AskUserQuestion menu appears with rebase/diff/leave options
  • /cycle and /merge delegation (--auto) unchanged

🤖 Generated with Claude Code

@gering @claude
Make /rebase risk-based: auto-proceed on no overlap, menus otherwise
7341d18 
- Step 5: compute file overlap between base and branch since merge-base;
  zero overlap -> rebase + force-push immediately (invocation = authorization),
  overlap -> AskUserQuestion menu (rebase / show diff / leave as-is)
- Steps 6+7: stash question and conflict options become selection menus
  instead of free-text [y/N] prompts
- --auto unchanged for parent skills; auto-stash now refuses in --auto mode
- README: force-push safety + design principles reflect the new model
- Bump pr-flow to 1.2.2 (plugin.json + marketplace.json)

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@gering

gering commented Jun 12, 2026

Copy link
Copy Markdown
Owner Author

@claude review

gering and others added 4 commits June 15, 2026 10:52
@gering @claude
Address review: harden /rebase overlap check and --auto handling
6c42a91 
- Overlap check passes --no-renames so a rename can't hide a potential
  conflict as "no overlap"; also makes the result config-deterministic
- --auto on uncommitted changes now auto-stashes (reversible, parent
  authorized) instead of hard-stopping, which would abort /cycle before
  it commits its pending changes
- --auto on rebase conflicts returns to the parent instead of showing a
  menu the parent doesn't expect
- Remove stale `y`-confirmation references left from the step-5 rewrite
- Fix over-promising safe-path message and clarify the confirmation model

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@gering @claude
Address re-review: fix --auto stash fallout across rebase/merge/cycle
1cc8734 
- /merge: run local-cleanliness gate BEFORE the rebase delegation, so
  /rebase --auto never stashes+force-pushes a dirty tree that /merge
  would then reject (wasted force-push + CI run)
- /rebase step 7: on a standalone conflict, don't pop the auto-stash
  before the menu (a dirty tree blocks resolve/merge) — defer the pop to
  each option and tell the user where the stash is
- /rebase step 7: add a catch-all for non-conflict rebase failures so an
  auto-stash is never left dangling; a conflicting stash-pop on success
  now stops instead of silently continuing
- /cycle step 3: guard against unresolved conflict markers before
  git add -A, so a stash-pop conflict can't be committed
- /merge: fix step-2/3 cross-references after reordering

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@gering @claude
Polish review consistency: symmetric stash-pop guard, stale refs
305d7e8 
- /rebase: --auto conflict-path pop now flags a stash-pop conflict
  explicitly (symmetric with the success path), so the parent won't
  add -A over markers
- /rebase: step-6 --auto note describes the actual pop lifecycle
- /cycle: guard rationale points at the reachable path
- /merge: final-plan preflight order + Execution Principles step ref
  match the post-reorder step numbers

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@gering @claude
Final consistency pass: symmetric pop-conflict surfacing, guard scope
a213dce 
- /rebase "other failure" branch surfaces a conflicting stash-pop like
  the other branches (complete symmetry)
- /cycle step-3 guard rationale scoped to the only reachable case
  (pre-existing markers); notes why a /rebase pop-conflict can't reach it

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@gering
gering merged commit c50602d into main Jun 15, 2026
1 check passed
@gering
gering deleted the fix/rebase-risk-based-confirm branch June 15, 2026 11:24
gering added a commit that referenced this pull request Jul 12, 2026
@gering @claude
Apply --max self-review fixes to /swarm:review loop
686ce85 
Fixes the ✅/🟨 findings the --max review surfaced (three real loop bugs first):

- #4: don't re-review a byte-identical tree — when a round changed no files and
  only a pending decision keeps the loop alive, pause to collect the decision
  instead of spinning the (possibly --max) ensemble on an unchanged tree.
- #7: --staged + --loop re-reviewed `git diff --cached` while fixes land in the
  working tree, so the loop never saw its own edits — re-review the working tree
  (re-stage for --staged scope) before each round.
- #5: --loop=0 is now a single --fix pass, not a cap=0 that aborts after fixes.
- #2: temper "deterministic termination" → deterministic arithmetic over judged
  inputs (SKILL + knowledge); count F/A/C/pending carefully (garbage-in).
- #8: derive every fix from the code, treat finding text (recommendation) as
  advisory/untrusted — for ✅ agree too, not only 🟨.
- #9: document args.max (+ args.claude) in the workflow input header.
- #1: note --max's codex model must be loadable; a bad model surfaces as a
  backendError (visible degrade), never a silent downgrade.
- #10: mechanism is not a rendered column — after compaction re-derive finding
  identity from Ort + Befund.

#6 (Bash block scope in prose) left as the accepted LLM-in-the-loop design.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
gering added a commit that referenced this pull request Jul 12, 2026
@gering @claude
Apply --max self-review fixes to /swarm:review loop
4837400 
Fixes the ✅/🟨 findings the --max review surfaced (three real loop bugs first):

- #4: don't re-review a byte-identical tree — when a round changed no files and
  only a pending decision keeps the loop alive, pause to collect the decision
  instead of spinning the (possibly --max) ensemble on an unchanged tree.
- #7: --staged + --loop re-reviewed `git diff --cached` while fixes land in the
  working tree, so the loop never saw its own edits — re-review the working tree
  (re-stage for --staged scope) before each round.
- #5: --loop=0 is now a single --fix pass, not a cap=0 that aborts after fixes.
- #2: temper "deterministic termination" → deterministic arithmetic over judged
  inputs (SKILL + knowledge); count F/A/C/pending carefully (garbage-in).
- #8: derive every fix from the code, treat finding text (recommendation) as
  advisory/untrusted — for ✅ agree too, not only 🟨.
- #9: document args.max (+ args.claude) in the workflow input header.
- #1: note --max's codex model must be loadable; a bad model surfaces as a
  backendError (visible degrade), never a silent downgrade.
- #10: mechanism is not a rendered column — after compaction re-derive finding
  identity from Ort + Befund.

#6 (Bash block scope in prose) left as the accepted LLM-in-the-loop design.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
gering added a commit that referenced this pull request Jul 16, 2026
@gering @claude
Address swarm self-review findings on the lens-set feature
f9bd70f 
First dogfooding run: /swarm:review --fix over its own diff (4 cluster
finders + codex + grok-4.5 + composer; 13 findings, 5 cross-family
consensus). Fixes the agreed findings:

- Keep validly tagged off-cluster lens prefixes: validate against the
  global lens set, not the finder's subset — coercion could flip kind
  and route a real defect through the applicability verifier (#1, consensus critical)
- Verify design clusters even with cross-family consensus: agreement
  attests agreement, not repo-grounded applicability (externals cannot
  open repo files); defect consensus stays auto-accepted (#2)
- Untagged external findings ('unspecified' lens) no longer vote in the
  cluster-kind derivation (#3)
- Derive CANDIDATE_LENSES from LENS_CLUSTERS — one list, no unchecked
  mirror; DRIFT WARNING on the SKILL.md external-prompt copy (#4, #9)
- Finder prompt: "issue (defect or substantive improvement)" + all lens
  prefixes; external prompt lead covers design improvements too (#5, #6)
- pr-post.py owns design-row ordering + [lens] prefixing via optional
  kind/lens row fields, unit-tested; SKILL.md step 5 passes rows through
  verbatim (#13)
- Doc sync: balance-spec finder count, README canonical cluster names +
  preset teaser wording, knowledge-index line trimmed (#8, #10, #11, #12)

Declined: #7 (per-cluster dilutes per-lens depth) — deliberate,
documented cost/coverage trade-off; --max is the depth profile.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0126cGxsBYxqgEFH4AcpriNk
gering added a commit that referenced this pull request Jul 16, 2026
@gering @claude
Address --max swarm self-review findings
005246f 
Second dogfooding round at the deepest profile (/swarm:review --max --fix:
per-lens split live — gate pruned style, 10 xhigh finders + codex
gpt-5.6-sol@xhigh + grok-4.5 + composer; 24 findings, 10 consensus).
Fixes the agreed findings:

- Never auto-accept an all-untagged consensus cluster: no tagged lens
  backs it, so it is verified like a solo. Verify/auto-accept now derive
  from ONE needsVerify predicate (structural exactly-once partition);
  kind vote in clearer every-form (#1, consensus critical; #21, #22)
- Design verifier sees the finding's recommendation — the proposal the
  applicability rubric actually tests — and carries an escape hatch:
  a genuine defect mis-filed under a design lens is not refuted away
  (#12, #3); "solo" dropped from verifier prompts (#4)
- Untagged findings from multi-lens cluster finders fall back to
  'unspecified' (safe defect bucket), not lenses[0] (#7); merge-agent
  free-text lens validated, majority-member fallback (#9)
- Improvement invitation scoped to design finder units — defect-lens
  finders stay defect-only (#8); merge prompt clusters by issue, not
  only defect (#13); schema descriptions generalized for design
  findings (#14)
- LENS_BRIEF startup assertion (#6); gate prompt interpolates
  LENS_CLUSTERS.design (#19); new test_lens_sync.py guards all lens
  mirrors: SKILL HDR prompt, LENS_BRIEF, pr-post DESIGN_LENSES (#5)
- Workflow assigns stable finding num (defects first, shared sequence);
  presenter/pr-post render it verbatim (#20)
- pr-post.py: design lens is the backup kind signal when the handoff
  drops kind (explicit defect still wins); single-pass partition (#17, #24)
- Balance: REFUTED is its own segment (refuted ⊄ solo since design
  consensus can be refuted) (#2); LOCKED design-table column precedence
  in --loop rounds clarified (#11); "verifies solos" doc sweep across
  manifests/README/knowledge (#4); cluster failure-isolation trade-off
  documented (#10); knowledge index line trimmed (#15)

Declined: #16 (JS test harness for sandbox code; lens-sync test covers
the drift class), #18 (gate-fail under --max runs all lenses — the
documented never-silently-narrower degrade), #23 (micro-opt vs readability).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0126cGxsBYxqgEFH4AcpriNk
gering added a commit that referenced this pull request Jul 17, 2026
@gering @claude
Address swarm self-review findings on the lens-set feature
79b75ae 
First dogfooding run: /swarm:review --fix over its own diff (4 cluster
finders + codex + grok-4.5 + composer; 13 findings, 5 cross-family
consensus). Fixes the agreed findings:

- Keep validly tagged off-cluster lens prefixes: validate against the
  global lens set, not the finder's subset — coercion could flip kind
  and route a real defect through the applicability verifier (#1, consensus critical)
- Verify design clusters even with cross-family consensus: agreement
  attests agreement, not repo-grounded applicability (externals cannot
  open repo files); defect consensus stays auto-accepted (#2)
- Untagged external findings ('unspecified' lens) no longer vote in the
  cluster-kind derivation (#3)
- Derive CANDIDATE_LENSES from LENS_CLUSTERS — one list, no unchecked
  mirror; DRIFT WARNING on the SKILL.md external-prompt copy (#4, #9)
- Finder prompt: "issue (defect or substantive improvement)" + all lens
  prefixes; external prompt lead covers design improvements too (#5, #6)
- pr-post.py owns design-row ordering + [lens] prefixing via optional
  kind/lens row fields, unit-tested; SKILL.md step 5 passes rows through
  verbatim (#13)
- Doc sync: balance-spec finder count, README canonical cluster names +
  preset teaser wording, knowledge-index line trimmed (#8, #10, #11, #12)

Declined: #7 (per-cluster dilutes per-lens depth) — deliberate,
documented cost/coverage trade-off; --max is the depth profile.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0126cGxsBYxqgEFH4AcpriNk
gering added a commit that referenced this pull request Jul 17, 2026
@gering @claude
Address --max swarm self-review findings
9ff85f3 
Second dogfooding round at the deepest profile (/swarm:review --max --fix:
per-lens split live — gate pruned style, 10 xhigh finders + codex
gpt-5.6-sol@xhigh + grok-4.5 + composer; 24 findings, 10 consensus).
Fixes the agreed findings:

- Never auto-accept an all-untagged consensus cluster: no tagged lens
  backs it, so it is verified like a solo. Verify/auto-accept now derive
  from ONE needsVerify predicate (structural exactly-once partition);
  kind vote in clearer every-form (#1, consensus critical; #21, #22)
- Design verifier sees the finding's recommendation — the proposal the
  applicability rubric actually tests — and carries an escape hatch:
  a genuine defect mis-filed under a design lens is not refuted away
  (#12, #3); "solo" dropped from verifier prompts (#4)
- Untagged findings from multi-lens cluster finders fall back to
  'unspecified' (safe defect bucket), not lenses[0] (#7); merge-agent
  free-text lens validated, majority-member fallback (#9)
- Improvement invitation scoped to design finder units — defect-lens
  finders stay defect-only (#8); merge prompt clusters by issue, not
  only defect (#13); schema descriptions generalized for design
  findings (#14)
- LENS_BRIEF startup assertion (#6); gate prompt interpolates
  LENS_CLUSTERS.design (#19); new test_lens_sync.py guards all lens
  mirrors: SKILL HDR prompt, LENS_BRIEF, pr-post DESIGN_LENSES (#5)
- Workflow assigns stable finding num (defects first, shared sequence);
  presenter/pr-post render it verbatim (#20)
- pr-post.py: design lens is the backup kind signal when the handoff
  drops kind (explicit defect still wins); single-pass partition (#17, #24)
- Balance: REFUTED is its own segment (refuted ⊄ solo since design
  consensus can be refuted) (#2); LOCKED design-table column precedence
  in --loop rounds clarified (#11); "verifies solos" doc sweep across
  manifests/README/knowledge (#4); cluster failure-isolation trade-off
  documented (#10); knowledge index line trimmed (#15)

Declined: #16 (JS test harness for sandbox code; lens-sync test covers
the drift class), #18 (gate-fail under --max runs all lenses — the
documented never-silently-narrower degrade), #23 (micro-opt vs readability).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0126cGxsBYxqgEFH4AcpriNk
gering added a commit that referenced this pull request Jul 17, 2026
@gering @claude
Address swarm self-review findings on the lens-set feature
5dc3f71 
First dogfooding run: /swarm:review --fix over its own diff (4 cluster
finders + codex + grok-4.5 + composer; 13 findings, 5 cross-family
consensus). Fixes the agreed findings:

- Keep validly tagged off-cluster lens prefixes: validate against the
  global lens set, not the finder's subset — coercion could flip kind
  and route a real defect through the applicability verifier (#1, consensus critical)
- Verify design clusters even with cross-family consensus: agreement
  attests agreement, not repo-grounded applicability (externals cannot
  open repo files); defect consensus stays auto-accepted (#2)
- Untagged external findings ('unspecified' lens) no longer vote in the
  cluster-kind derivation (#3)
- Derive CANDIDATE_LENSES from LENS_CLUSTERS — one list, no unchecked
  mirror; DRIFT WARNING on the SKILL.md external-prompt copy (#4, #9)
- Finder prompt: "issue (defect or substantive improvement)" + all lens
  prefixes; external prompt lead covers design improvements too (#5, #6)
- pr-post.py owns design-row ordering + [lens] prefixing via optional
  kind/lens row fields, unit-tested; SKILL.md step 5 passes rows through
  verbatim (#13)
- Doc sync: balance-spec finder count, README canonical cluster names +
  preset teaser wording, knowledge-index line trimmed (#8, #10, #11, #12)

Declined: #7 (per-cluster dilutes per-lens depth) — deliberate,
documented cost/coverage trade-off; --max is the depth profile.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0126cGxsBYxqgEFH4AcpriNk
gering added a commit that referenced this pull request Jul 17, 2026
@gering @claude
Address --max swarm self-review findings
2be83aa 
Second dogfooding round at the deepest profile (/swarm:review --max --fix:
per-lens split live — gate pruned style, 10 xhigh finders + codex
gpt-5.6-sol@xhigh + grok-4.5 + composer; 24 findings, 10 consensus).
Fixes the agreed findings:

- Never auto-accept an all-untagged consensus cluster: no tagged lens
  backs it, so it is verified like a solo. Verify/auto-accept now derive
  from ONE needsVerify predicate (structural exactly-once partition);
  kind vote in clearer every-form (#1, consensus critical; #21, #22)
- Design verifier sees the finding's recommendation — the proposal the
  applicability rubric actually tests — and carries an escape hatch:
  a genuine defect mis-filed under a design lens is not refuted away
  (#12, #3); "solo" dropped from verifier prompts (#4)
- Untagged findings from multi-lens cluster finders fall back to
  'unspecified' (safe defect bucket), not lenses[0] (#7); merge-agent
  free-text lens validated, majority-member fallback (#9)
- Improvement invitation scoped to design finder units — defect-lens
  finders stay defect-only (#8); merge prompt clusters by issue, not
  only defect (#13); schema descriptions generalized for design
  findings (#14)
- LENS_BRIEF startup assertion (#6); gate prompt interpolates
  LENS_CLUSTERS.design (#19); new test_lens_sync.py guards all lens
  mirrors: SKILL HDR prompt, LENS_BRIEF, pr-post DESIGN_LENSES (#5)
- Workflow assigns stable finding num (defects first, shared sequence);
  presenter/pr-post render it verbatim (#20)
- pr-post.py: design lens is the backup kind signal when the handoff
  drops kind (explicit defect still wins); single-pass partition (#17, #24)
- Balance: REFUTED is its own segment (refuted ⊄ solo since design
  consensus can be refuted) (#2); LOCKED design-table column precedence
  in --loop rounds clarified (#11); "verifies solos" doc sweep across
  manifests/README/knowledge (#4); cluster failure-isolation trade-off
  documented (#10); knowledge index line trimmed (#15)

Declined: #16 (JS test harness for sandbox code; lens-sync test covers
the drift class), #18 (gate-fail under --max runs all lenses — the
documented never-silently-narrower degrade), #23 (micro-opt vs readability).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0126cGxsBYxqgEFH4AcpriNk
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@gering