refactor(prose): dump/read folds — subtree gets replace per-topic loops (D7, stage 5) - #509
Merged
Conversation
This was referenced Jul 22, 2026
Merged
Merged
leeovery
force-pushed
the
feat/set-grammar-uniform
branch
from
July 23, 2026 11:04
014be35 to
49f3fb9
Compare
leeovery
force-pushed
the
feat/batch-read-folds
branch
from
July 23, 2026 11:04
37e7791 to
5f71091
Compare
leeovery
force-pushed
the
feat/set-grammar-uniform
branch
from
July 23, 2026 12:29
49f3fb9 to
5da39dc
Compare
leeovery
force-pushed
the
feat/batch-read-folds
branch
from
July 23, 2026 12:29
5f71091 to
13aa798
Compare
leeovery
force-pushed
the
feat/set-grammar-uniform
branch
from
July 23, 2026 12:30
5da39dc to
1d3e4df
Compare
leeovery
force-pushed
the
feat/batch-read-folds
branch
from
July 23, 2026 12:30
13aa798 to
286a34d
Compare
leeovery
force-pushed
the
feat/set-grammar-uniform
branch
from
July 23, 2026 12:31
1d3e4df to
7d3919b
Compare
leeovery
force-pushed
the
feat/batch-read-folds
branch
from
July 23, 2026 12:31
286a34d to
c3b9771
Compare
…ps (D7, stage 5)
Zero engine code again — the whole-subtree get was always there.
- sequence-discovery-map: 2×N summary/description gets → one
{wu}.discovery subtree read.
- read-plans: per-plan format/external_id gets (2×N) → one
{wu}.planning subtree read; the per-plan loop keeps only judgment.
- brief-synthesis: lifecycle cleanup collects orphans across
split/merge/drop into one rm -f + one apply (brief_path deletes);
propagation reads research+discussion subtrees once (2 calls total,
any N) and lands every reconcile_needed flag in one apply.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
leeovery
force-pushed
the
feat/batch-read-folds
branch
from
July 23, 2026 12:32
c3b9771 to
413c3ae
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stage 5 of the batching programme (design log: #504; base: #508). The dump/read-fold class — and like stage 3, C5 keeps it at zero engine code:
manifest get <wu>.<phase>has always returned the whole subtree; the prose just never leaned on it.What
summary/descriptiongets become one{wu}.discoverysubtree read (the dump's rows deliberately stay lean —description_presentonly — so the subtree get is the right channel for the full text).format/external_idgets become one{wu}.planningsubtree read before the loop; the per-plan loop keeps only the judgment work.rm -fnaming all files, oneapplydeleting allbrief_pathpointers (the loud-on-absent guard preserved: only topics that carried the pointer enter the payload).reconcile_neededflags, lands them in oneapply.Test plan
Prose-only:
npm test1556/1556, conventions lint 27/27, typecheck clean.🤖 Generated with Claude Code
Stack