revdiff-planning vs planning@cc-thingz: both hook ExitPlanMode, different feature sets — intended relationship? #269
Replies: 1 comment
|
running both isn't something I designed for. Both plugins hook the history is the other way round from what it looks like.
one more thing about running both. If you review clean in the revdiff-planning pane but annotate in planning's, rolling diff missing from fixed the docs in #362 - README and the site said this plugin "does not conflict with other planning plugins", which was wrong for exactly your setup. |
Uh oh!
There was an error while loading. Please reload this page.
I run both
revdiff-planning@revdiff(0.3.8) andplanning@umputun-cc-thingz(3.8.4). Both register aPreToolUsehook onExitPlanMode, so both fire on every plan review.Feature diff I found comparing the two
launch-plan-review.sh/plan-review-hook.pypairs:revdiff-planninghas rolling/compare-mode review — a<!-- previous revision: ... -->marker + snapshot chain, so round 2+ only shows the diff since the last round (--compare-old/--compare-new --collapsed).planning@cc-thingzhas none of that (always--only, full plan every round), but has a native agterm branch (agtermctl session overlay open ... --block) and a$EDITORfallback viaplan-annotate.py, plusPLANNING_DISABLE_REVDIFFto bail out cleanly (useful for/remote-control).Looks like
planningmight have forked/evolved fromrevdiff-planningand picked up agterm + editor-fallback along the way, but dropped rolling-diff. With both installed, Claude Code runs both hooks on everyExitPlanMode(per the hooks doc,PreToolUseverdicts merge as most-restrictive-wins) — so right now I get two review passes back to back, which is redundant.Is coexistence intended, or should one supersede the other? If intended, is
PLANNING_DISABLE_REVDIFF=1(silencingplanning's own review, lettingrevdiff-planningbe the sole reviewer) the recommended way to combine them? If not — is rolling-diff missing fromplanningan oversight, or dropped on purpose?All reactions