Repository navigation
style: satisfy prettier across the repo - #113
Merged
Merged
Conversation
Mechanical reformat only — prettier --write over the 11 files the CI "Format check" step reports. No logic changes: the sole token-level deltas are semicolon separators inside inline TypeScript type literals becoming newlines when prettier reflows them (semi: false), plus markdown reflow and quote normalisation. Ten of the eleven were already failing on main before PR #112. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Env.ts returns process.env.PRIVY_APP_ID unvalidated whenever it is set, so presence alone satisfies it. Under vitest NODE_ENV is 'test', so the non-production branch threw at module-import time and five suites failed during collection, before any in-test vi.stubEnv could run. Latent on main, not new: Format check is step 3 and aborted the job before tests ever ran, so this red has never surfaced. Making format pass exposed it. Plain env, not a secret — no test contacts Privy. Co-Authored-By: Claude Opus 5 <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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
prettier --writeover the 11 files the CIverifyjob's Format check step reports. Formatting only — no logic changes.Why this is a separate PR, not a push to #112
The task was to fix this on
chore/drop-root-skills-link-layer(PR #112). #112 was merged to main by @hishboy at 2026-08-26T13:22:54Z (squash →5d737b3) while this work was being picked up, and the branch and its worktree are gone. So the same fix lands here, off currentmain, instead.Scope
mainhas failed CI on its last 5 commits, not just #112. Of the 11 files, onlyAGENTS.mdwas ever in #112's diff — the other 10 were already failing on main beforehand:Per hish's ruling — nothing red merges, and main does not rot — all 11 are fixed here.
Proof it is formatting only
Every changed file was compared old-vs-new with all whitespace stripped and quotes normalised. Six files are byte-identical under that normalisation. The five that are not:
3 markdown files — list-marker, table-padding, and hard-break reflow (adds/removes real characters by design).
HyperliquidService.test.ts(2 tokens),TrailingStopService.test.ts(8 tokens) — the only code-token change in the entire diff is;separators inside inline TypeScript type literals turning into newlines when prettier reflows them onto multiple lines, e.g.;and newline are interchangeable member separators in a TS type literal. No runtime or type-semantics effect. Driven by the repo's ownsemi: falseinprettier.config.mjs.No statement-terminating semicolon was removed anywhere —
grep -nE ';\s*$'finds zero such lines in either file at the base commit.Verification
bun run format:check— the exact CI command — is clean:Non-vacuity controls, because a formatter that silently checks nothing looks identical to a passing one:
git ls-filescross-check agrees);--list-differentreturns 0 files;--checkand exits 1, proving the instrument reads files rather than passing by default;prettier <file>→ exit 0, non-empty output) before any--write, so no blind write ran against an unparseable file.One quirk worth recording:
HyperliquidService.test.tsneeded two--writepasses to converge. Pass 1 left it still failing--check; pass 2 reached a fixed point, stable on a third check.Caveat on "green"
Format check is step 3 of 6 in
verify(install → format → lint → build → vitest → skills digest). Steps 4-6 have not run on main in 5 commits because format aborted the job first. This PR unblocks the job to reach lint; whether lint, build, vitest, and the skills digest then pass is unknown until CI runs here. Not asserting green beyond what CI reports.🤖 Generated with Claude Code
Second commit: placeholder
PRIVY_APP_IDfor the test stepFixing format exposed a second, latent red on main:
Format checkis step 3 and aborted the job before tests ever ran, somainhas never surfaced it. First run here:Format check✅Lint✅Build✅ then 22 failures, all one cause —Error: PRIVY_APP_ID is not set(5 collection-time suite failures + 17 test failures;6 failed | 36 passed (42)). ZeroSyntaxError/Unexpected token/Cannot findanywhere, i.e. not the reformat.Presence, not identity — established from the source, not by trial.
src/common/Env.ts:37:Any non-undefined
process.env.PRIVY_APP_IDshort-circuits the??and is returned unvalidated —ensureStringis only reachable when the variable is unset. Under vitestNODE_ENVis'test', soIS_PRODUCTIONis false and that branch threw at module-import time, which is why the suites died during collection, before any in-testvi.stubEnvcould run.Nothing in the suite contacts Privy: the tests that do exercise it already stub their own fakes (
'privy-app-id','test-privy-app-id'inEnv.test.ts,LoginService.test.ts,WalletService.test.ts) and pass today. So the fix is one plainenv:line on the test step — deliberately not a secret, since a real app id would be both unnecessary and a live credential sitting in CI.CI result — fully green
Run
32974881241, every step, none skipped:gh pr checks 113→verify pass. Test files went6 failed | 36 passed→42 passed, and the digest step ran for the first time rather than being skipped behind an earlier failure.