fix(ci): remove duplicate Python block dangling outside heredoc#13
Merged
Conversation
The rsr-antipattern workflow had two PYEOF tokens in succession but only one opening python3 heredoc, leaving the second block (lines 140-215) as bash commands. Bash saw as a command and exited 127, failing every CI run on main since the check was added. The first Python block already does the full TypeScript exemption check (honouring the universal allowlist + .claude/CLAUDE.md table), so the duplicated second block is just removed. No behaviour change beyond the workflow actually running to completion.
3 tasks
hyperpolymath
added a commit
to hyperpolymath/standards
that referenced
this pull request
May 11, 2026
## Summary Two related CI fixes — both unblocking workflows that have been red on `main` for weeks. ### 1. `.github/workflows/rsr-antipattern.yml` — heredoc syntax bug The "Check for TypeScript" step had two `PYEOF` tokens in succession but only one opening `python3 << 'PYEOF'`, leaving the second Python block dangling as bash commands. Bash saw `BUILTIN_GLOBS = [...]` as a command and exited 127. Every commit to `main` since the check was added has hit this. Same bug was in `hyperpolymath/my-lang` (see [hyperpolymath/my-lang#13](hyperpolymath/my-lang#13)). The first Python block already does the full TypeScript exemption check honouring the universal allowlist and the parsed `.claude/CLAUDE.md` table. The duplicated second block is removed. ### 2. `.github/workflows/language-policy.yml` — duplicate TS check with no allowlist The "Check for TypeScript files" step did a blunt `find . -name "*.ts"` filter with no allowlist beyond `node_modules` and `*.d.ts`. False-positived on legitimate bridge files like `lol/test/vitest.config.ts`, which lives under the `**/test/**` universal-allowlist pattern that `rsr-antipattern.yml` honours. Two checks for the same banned-language family is also a single-source-of-truth violation. This PR removes the duplicate from `language-policy.yml`; `rsr-antipattern.yml` remains the canonical TypeScript-detection check. All other `language-policy.yml` steps (ReScript, Go, Python, V-lang, ATS2, Java/Kotlin, Swift, Flutter/Dart, Makefiles, package.json runtime deps) stay — they remain the single source of truth for those languages. ## What this PR fixes vs. what stays red **Fixed by this PR:** - `antipattern-check` no longer exits 127 on the heredoc syntax bug. It now runs to completion. - `Check for Banned Languages / Check for TypeScript files` no longer false-positives on `lol/test/vitest.config.ts`. **Surfaces real pre-existing migration debt (out of scope for this PR):** - `antipattern-check / Check for ReScript` flags ~30 real `.res` files under `rhodium-standard-repositories/satellites/{consent-aware-http,cccp/.../7-tentacles}/`. These were previously hidden behind the heredoc crash — the check never reached this step before. The fix here unblocks the workflow's ability to surface them; the actual migration is its own multi-PR effort. - `Check for Banned Languages / Check for ReScript files` (in `language-policy.yml`) also flags the same `.res` files. Same observation. - `Hypatia Neurosymbolic Analysis (Dogfooding)` — `hypatia-cli.sh scan .` exits 1 with no diagnostic across multiple repos. Filed at [hyperpolymath/hypatia#213](hyperpolymath/hypatia#213). So this PR doesn't turn the standards CI fully green — that would need the ReScript migration and the Hypatia scanner fix. But it does: 1. Unblock the antipattern check's *ability to run*. 2. Eliminate a false-positive that was masking real signal. 3. Halve the spurious red-X load on every PR. ## Why now Both bugs surfaced while merging the affinescript#64 port work in #46. Pre-existing — unrelated to that port — but kept its merge state at `BLOCKED` until force-merged with `--admin`. Fixing now so future port PRs have less friction. ## Test plan - [x] CI on this PR shows `antipattern-check` running to completion (not exit 127). - [x] `Check for Banned Languages / Check for TypeScript files` step passes (no spurious vitest.config.ts hit). - [ ] Verified by sight: the only remaining failures on this PR are pre-existing — ReScript migration debt and the Hypatia scanner bug — not introduced or worsened here. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
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
The
RSR Anti-Pattern Checkworkflow has been red on every commit tomainsince the check was first added (5+ runs visible in the workflow history, last green: none). Root cause: the "Check for TypeScript" step has twoPYEOFtokens but only one openingpython3 << 'PYEOF', so the second Python block dangles as bash commands. Bash interpretsBUILTIN_GLOBS = [...]etc. as commands and exits 127.The first Python block already does the full TypeScript exemption check (universal allowlist + parsed
.claude/CLAUDE.mdexemption table). The duplicated second block is redundant — removing it makes the workflow actually run to completion.Verification
Confirmed by comparing to
affinescript's equivalent workflow (which has only 2PYEOFtokens — one opener, one closer — and consistently passes).Why now
Discovered while merging affinescript#64-related port work. Pre-existing — unrelated to that port — but kept the merge in a
BLOCKEDstate until force-merged with--admin. Fixing now so future port PRs don't have the same friction.Test plan
🤖 Generated with Claude Code