Keep comms/ and research/ out of the built package - #235
Merged
Conversation
`R CMD build` ships every top-level directory not named in `.Rbuildignore`,
so `pak::pak("NewGraphEnvironment/link")` was putting internal cross-repo
coordination notes and research working files into the installer's library.
`R CMD check` reports this only as a NOTE, and `.gitignore` does not cover it.
Verified against the tarball rather than the config: 37 `comms/` + `research/`
entries before, 0 after.
`planning`, `dev`, `.claude` and `CLAUDE.md` were already excluded - these two
were added later and missed. Convention and an audit loop for the other repos
are in NewGraphEnvironment/soul#60.
Co-Authored-By: Claude <[email protected]>
Claude-Session: https://claude.ai/code/session_01EPc4Xkk81mE4JHuP5A1SnS
…ind) * main: (43 commits) Pack components by finish time, not segment count Fix two pre-flight gates the first real pilot exposed (v0.47.1) Fail loudly when a fresh cypher never accepts a connection Assert per-host drainage closure and DS-first order (#246) code-check round 5 clean; sweep the last two unsafe greps (#246) code-check round 4: the empty-bucket guards could never run (#246) code-check round 3: stop the completeness gate destroying good work (#246) code-check round 2: fix a blocker inside round 1's fix (#246) code-check round 1: fix the TSV seam and four fragile spots (#246) Phase 4: RUNBOOK section 8d, NEWS, v0.47.0 Phase 3: derive the host buckets instead of asserting them (#246) Phase 2: pre-flight gates for the study-area run (#246, absorbs #183) Phase 1: make fresh a real dependency; fix the prep sentinel (#246) Initialize PWF baseline for #246 CLAUDE.md: sync version header to v0.46.0 (memory-audit, soul#47) Update CLAUDE.md with soul conventions Archive planning files for issue #227 Wrap long lines in the guard message builder (#227) Phase 5 (#227): RUNBOOK 8c, guard-note docs, CLAUDE.md status, release v0.46.0 Phases 3-4 (#227): wire the guard into the runners, add the post-condition ...
Owner
Author
|
Updated the branch before merging: it was 55 commits behind main and its Merged main in; the branch's delta vs main is now exactly the 2 lines it describes, and settings match main. Re-measured the tarball rather than trusting the month-old table —
Same result, current evidence. (Corrected: an earlier revision of this comment mistyped the |
This was referenced Aug 31, 2026
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.
Summary
R CMD buildships every top-level directory not named in.Rbuildignore.linkis an installable package, sopak::pak("NewGraphEnvironment/link")was putting internal cross-repo coordination notes (comms/) and research working files (research/) into the installer's library.R CMD checkreports this only as a NOTE, and.gitignoredoes not cover it.Verification
Against the built tarball rather than the config, since the
.Rbuildignoreregex is easy to get subtly wrong:comms/research/Why it was missed
planning,dev,.claudeandCLAUDE.mdwere already excluded.comms/andresearch/were added to the repo later and never added to the ignore list — the gap opens over time rather than at scaffold.This matters for the three-layer repo split:
comms/is internal by definition, so a package that flips public while shipping it leaks exactly what the flip was meant to purge.Notes
Only
.Rbuildignoreis touched. The branch was cut from a working tree with 226 unrelated in-flight changes; none of them are staged or committed here.Convention entry and an audit loop covering the other repos: NewGraphEnvironment/soul#60
🤖 Generated with Claude Code
https://claude.ai/code/session_01EPc4Xkk81mE4JHuP5A1SnS