proofpage report

commit
550dfe6c517156fbf3117a04ac5b2fd7fa8b96aa
branch
main
working tree
DIRTY - uncommitted changes were present when these checks ran
node
v24.15.0
platform
win32 x64 (10.0.26200)
measured at
2026-09-05T11:48:34.205Z (UTC)
checks discovered via
auto-detected from package.json scripts

2 of 2 checks passed.

test

PASS
command
npm run test
exit code
0
duration
8.35s
measured at
2026-09-05T11:48:34.275Z (UTC)
discovered via
auto-detected from package.json
parsed result
node:test summary - pass 54, fail 0, total 54
output (last 40 of 67 lines, truncated)
✔ --check FAILS when there is no print stylesheet (0.1021ms)
✔ --check FAILS on unbalanced <details> (0.1247ms)
✔ --check FAILS on unbalanced <pre> (0.2267ms)
✔ --check FAILS on a collapsed <details> (missing the open attribute) (0.1646ms)
✔ --check FAILS on a credential-shaped string (0.1801ms)
✔ --check FAILS on missing doctype (0.1122ms)
✔ renderProof output always passes --check (the renderer only ever produces honest HTML) (0.6021ms)
✔ a check that did not run shows no exit code number and no parsed count (0.7792ms)
✔ a check with unparseable output renders "unparsed" and the reason, never a fabricated 0 (0.2784ms)
✔ a dirty working tree surfaces a visible warning (0.271ms)
✔ when git is unavailable, the header says so rather than showing a fabricated commit (0.1971ms)
✔ renderProof redacts a credential-shaped string even if it arrives already embedded in proof data (defense in depth) (0.2965ms)
✔ discoverChecks reads proofpage.json when present, and marks entries as not auto (1.3308ms)
✔ discoverChecks auto-detects from package.json scripts when there is no proofpage.json (3.0781ms)
✔ discoverChecks reports "none" when there is nothing to discover (0.7295ms)
✔ runOneCheck actually executes the command and captures a real passing exit code (679.2467ms)
✔ runOneCheck actually executes the command and captures a real non-zero exit code (592.3617ms)
✔ runOneCheck parses a real node:test summary from a real `node --test` run (683.5404ms)
✔ runOneCheck records unparsed, never a fabricated count, when the output shape is unknown (94.2215ms)
✔ runOneCheck marks a check that could not be spawned as ran:false, never as a fake pass (1.1697ms)
✔ runOneCheck strips NODE_TEST_CONTEXT so a nested `node --test` check is never silently skipped by node's own recursion guard (741.2059ms)
✔ runOneCheck redacts credential-shaped strings from captured output before they reach proof.json (115.6035ms)
✔ parseOutput reads a real tsc-shaped "Found N errors" line (0.1829ms)
✔ parseOutput also recognizes the classic TAP-reporter shape ("# pass N" / "# fail N") (0.2717ms)
✔ parseOutput handles the singular "Found 1 error." form (0.0624ms)
✔ parseOutput does not guess a count when nothing matches a known shape (0.0417ms)
✔ getGitInfo reports available:false outside of any git repository (21.7981ms)
✔ getGitInfo detects a dirty working tree honestly, against a real isolated git repo (362.9867ms)
✔ run() writes proof.json to disk and computes a real summary from real exit codes (635.1712ms)
✔ run() against the fixture repo computes an honest summary across pass/fail/parsed/unparsed checks (1813.7013ms)
✔ run() records env and git fields without fabricating them (1879.1211ms)
ℹ tests 54
ℹ suites 0
ℹ pass 54
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 7868.746

lint

PASS
command
npm run lint
exit code
0
duration
1.25s
measured at
2026-09-05T11:48:42.627Z (UTC)
discovered via
auto-detected from package.json
parsed result
unparsed - output did not match a known shape (node:test summary or tsc "Found N errors")
output (6 lines)
> proofpage@0.1.0 lint
> node scripts/lint.mjs

10/10 file(s) passed syntax check