[bench] comments: trim change-narration, dedupe rationale, add graphify docs - #28
Merged
Conversation
Records the graph-first workflow for structural questions: query the graphify-out/ knowledge graph via query/path/explain before exploratory grep, read cited files before editing, refresh with --update after commits. Forbids `graphify claude install` here since it writes CLAUDE.md directly, bypassing governance. Includes the governance receipts for the edit. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XzAGxRQq2ts2MGciKc8btC
A comment audit found three removable categories concentrated in
chain.py, api.py, runner.py, and constitution.py: narration addressed
to a past reviewer ("behavior is unchanged", "NOT a rename", a
tombstone for a deleted function), the same rationale told in several
places (the frozen-array story four times in chain.py, the soft-
dependency story twice in api.py), and restatements of CLAUDE.md or
other modules. Each rationale now has one authoritative telling with
pointers elsewhere; every hazard-documenting comment stays.
Logic is provably untouched: all four files are AST-identical to the
previous commit with docstrings stripped. The pass also corrected a
stale claim in append_entry's docstring, which still said the function
updates ledger-meta.json; that stopped being true when the legacy
segment was frozen.
Includes the governance receipts for these edits, among them one VETO
on a scratchpad audit script whose except block swallowed errors,
rewritten and re-adjudicated as PASS.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XzAGxRQq2ts2MGciKc8btC
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
1 participant
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.
What
Two commits:
graphify claude installwrite path).ledger/chain.py,utils/api.py,pipeline/runner.py,pipeline/constitution.py: removes reviewer-addressed change-narration, deduplicates rationale told in multiple places down to one authoritative telling plus pointers, and drops restatements of CLAUDE.md. Net 46 comment lines removed; every hazard-documenting comment kept.Why
Prepping the repo for public readers: change-narration ("behavior is unchanged", tombstones for deleted functions) only means something to someone who saw the original diff, and duplicated rationale is a drift risk in files that are fed to the pipeline as evidence. The audit proved the risk is real:
append_entry's docstring still claimed it updatesledger-meta.json, which stopped being true when the legacy segment froze. Fixed here.Verification
python -m cli verify: chain intact, meta anchor verified.