* fix(robots-check): stop a leading BOM or an undecodable rule from reading as allowed
Two decoding edge cases left the robots gate failing open. Both were
flagged as follow-ups in #506.
A robots.txt saved with a UTF-8 byte-order mark decodes to a body that
starts with U+FEFF, so its first field read as U+FEFF + "user-agent". A
leading "User-agent: *" went unseen, every rule in that group was
dropped for want of an agent, and "Disallow: /" read as allow-all.
_groups() and is_robots_body() now skip one leading U+FEFF, as Google's
reference parser does, which covers every caller of allowed(). This
predates #506: the old code-page decode hid the line the same way.
Since #506 pinned curl's output to UTF-8 with errors='replace', a
non-conformant robots.txt saved in a legacy code page (cp1254,
ISO-8859-9) with raw non-ASCII bytes in a rule decodes them to U+FFFD.
That rule can never match and was silently skipped. The new
has_undecodable_rule() treats such a body as unreadable, so the gate
returns UNCONFIRMED (exit 1). The field is named with U+FFFD removed, so
a stray byte before "Disallow" cannot hide the line either. U+FFFD in a
comment or another field (Sitemap, Crawl-delay) stays harmless, and
valid UTF-8 non-ASCII rules decide as before.
Nine new tests in two FAIL-OPEN REGRESSION classes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs(changelog): record the robots_check BOM and U+FFFD fixes
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Can Mugan <217335573+canmugan@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
find_orphans() compared each line's left edge with the document margin:
an un-indented last line followed by an indented first line was an entry
header orphaned from its bullets. A list marker moves a line's left edge
without moving its text, so the rule misread two templates:
* one that merges each bullet's marker into its first line (#481): a
bullet wrapping across the break was reported as an orphaned header,
and a real orphaned header followed by such bullets was not reported;
* the stock CV, when its outer marker extracts as a line of its own: a
[Job Title] header left at the foot of a page, bullets overleaf, was
not reported (its text at x70.4 counts as indented), and a one-line
list item whose outer marker is set 1.3pt below it - so the marker
sorts last on the page - was reported as a split bullet although the
whole item is on that page.
Following discussion #481, the rule now reads text x:
* parse_pdf records where a line's text starts after leading
LIST_MARKERS words (Line.text_left, defaulted). The set is the agreed
Unicode bullets without the dashes (`• ‣ ▪ ● ·`): a bullet's second
line may open with an em dash, and the stock Honors item's second line
opens with an en dash. Other glyphs are text, so a line that merges
the stock outer marker (`○` on some setups, an emoji on others) keeps
the marker's x, as the left-edge rule read it;
* find_orphans compares the text x of a page's last printed line with
that of the next page's first printed line, and reports an orphaned
header when the latter is more than INDENT_PT further in. Bbox lines
whose tops are within SAME_PRINTED_LINE_PT (3pt) are one printed line:
pdftotext's rounded-yMin buckets split a line where a bold word sits
0.2pt off, or where a marker sits 1.3pt below its text. A line whose
text starts more than SET_IN_LIMIT_PT (60pt) past the margin is set
right or centred (a running header) and is skipped as the opener;
* a page ending on a marker with no text beside it keeps the lone-marker
report, with its condition and message unchanged.
Stock cv/main_example.tex, \vspace* of 0-600pt in 8pt steps before four
entries, 304 lualatex builds (and the same 304 with xelatex): this adds
15 reports, each a [Job Title] header at the foot of a page with its
bullets overleaf; the other 289 builds give byte-identical output. The
reported.
A two-line outer item whose marker sorts last and that wraps across the
break is no longer reported as a split bullet: it is read like the #481
wrapped bullet, which is expected clean.
Tests: TestOrphansReadTextX (13; 7 fixes and the wrapped two-line item
fail on master's logic, 5 pin behaviour master already had) and
TestTextStart (4). On Python 3.10-3.14 `python -W error::SyntaxWarning
-m unittest discover -s tests -t .` passes 520 tests (master 503);
lint_skills.py, check_framework_version.py and security_guards.py OK.
* fix(job_key): hash the URL alone so a re-listed non-Latin title keeps one key
* docs(changelog): record the job_key URL-identity fix
Adds the [Unreleased] / Fixed entry for #501, naming tools/job_key.py and
tests/test_job_key_url_identity.py.
The entry states the migration consequence in bold: an entry written before
this fix, for a non-Latin title on a portal without a numeric id, re-keys once
on the next scrape. It also names where those entries surface,
keys_not_matching_current_rule, and that --audit never rewrites them.
Verified that claim rather than asserting it. A seen_jobs.json holding one
entry under its pre-fix key for a Cyrillic-titled freehire URL reports
keys_not_matching_current_rule: 1 under the fixed audit. The same audit over a
real 884-entry seen_jobs.json reports 0, because those keys are the
company_title form that is_canonical() excludes from the drift check.
Rolls [Unreleased] into [1.7.2] - 2026-09-29 and moves the compare links.
Keeps an empty [Unreleased] heading above the release, per Keep a Changelog.
Thirty-two entries: seven added, one changed, twenty-four fixed.
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Titles are indexed as the source ATS wrote them and toResult passed them
through untouched, so `detail intern-fullstack-amp-ai-innovation-noise-edaawqq2`
printed `Intern - Fullstack & AI Innovation ` and the same title reached
`search` output. In a 93-result /scrape batch, 1 title carried `&` and
3 ended in whitespace; /scrape Step 4.75 reads an undecoded entity in a
title as a half-working parser.
Run the title through the existing decodeHtmlEntities (already used for
detail descriptions) and trim it; an empty or whitespace-only title still
falls back to "(untitled)". Three new parsing.test.ts cases pin it; all
three fail on the old pass-through.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
formatSalary printed only the currency and amounts, although freehire's
enrichment records the pay period. `detail software-developer-trainee-codifi-fra635ux`
printed `INR 300000–300000` while the API returns `salary_period: "year"`
for that posting, so a yearly salary screened against a monthly figure
read as twelve times its real monthly value.
Append the period when freehire records one (`INR 300000–300000/year`,
`INR 12000/month`); a record without a period prints exactly as before.
Two new parsing.test.ts cases pin it; both fail on the old formatter.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
_fetch() ran curl with text=True and no encoding, so Python decoded the
response with the Windows ANSI code page. On a byte that code page leaves
undefined (0x81 in cp1252 and cp1254, 0x9e in cp1254) subprocess's pipe
reader thread died, stdout came back None, rpartition() raised
AttributeError, and the gate printed "UNCONFIRMED (AttributeError)". That
fails closed, but it blocked the browser-header retry 09-web-research.md
permits on hosts such as tr.indeed.com (a UTF-8 CJK Disallow rule) and
kap.org.tr (a UTF-8 Turkish 404 page).
Pin encoding='utf-8', errors='replace' on the call (RFC 9309: robots.txt
is UTF-8), as verify_pdf.py and verify_layout.py already do. Decision
rules and the fail-closed handling of real read failures are unchanged.
The regression tests swap curl for a child process that prints the real
bytes, so the actual subprocess decode path runs with _fetch's own
kwargs, and pin cp1254 when no encoding is passed so a UTF-8 host cannot
pass the bug by luck. Both fail on master.
Co-authored-by: Can Mugan <217335573+canmugan@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
FindNonAsciiDateRangesTests' docstring (from #498) quotes the regex `\s*`
in a non-raw string. `\s` is an invalid escape sequence, so Python 3.12+
prints a SyntaxWarning for it (CPython gh-98401) when
tests/test_verify_pdf.py is compiled. It is in this repo's own CI log
for a45bbd1 on 3.12, 3.13 and 3.14, at tests/test_verify_pdf.py:75, while
the run stays green. Under `-W error::SyntaxWarning` the module fails to
import and the whole test_verify_pdf suite drops out of the run.
Escaped in the docstring. ToolsCompileWithoutWarnings (#480) only compiled
tools/*.py, which is why this got through; it now compiles tests/*.py too.
With the unescaped docstring it fails naming tests/test_verify_pdf.py, and
it passes with the fix.
Verification on Python 3.10, 3.11, 3.12, 3.13 and 3.14, from a clean
bytecode cache: `python -W error::SyntaxWarning -m unittest discover -s
tests -t .` OK (501 tests, the same count as master),
`tools/lint_skills.py`, `tools/check_framework_version.py`,
`tools/security_guards.py`.
* feat(verify_pdf): add --ascii-dates, the non-folding check for the documented date-range rule (#385)
05-cv-templates.md records a real Workday import that dropped a role's
end date because `2016--2024` reaches the text layer as 2016<U+2013>2024,
and tells Step 5d to confirm every entry's years are joined by an ASCII
hyphen. Nothing executed that. The natural probe, --contains "2016-2024",
failed identically for "en-dashed" and "absent" before #458 and passes on
both since, because the fold maps the en-dash back to "-" on purpose.
Add --ascii-dates: scan the raw text layer (never the fold) for a year
joined to any Unicode dash or the minus sign and fail naming each hit
with its line and code point. A year on either side of the dash is
enough (Mar 2016 - Jul 2016, 2016 - Present); a numeric range with no
year is not a date. The dump is still written before the check fails.
Wire it in: Step 5d's extraction command carries the flag and the
"Dates recognizable" item points at it, keeping the bare single-year
half as a read-through; the guide's "add this to the step 5d checks"
paragraph names the command (framework_version 1.4.4 -> 1.4.5); the
upstream-only CI assertion runs it on the stock CV.
Demonstrated on the stock template compiled with lualatex through both
extractors: a \cventry{2016--2024} build fails with U+2013 while
--contains "2016-2024" passes on it; the \cventry{2016-2024} build passes
both; the untouched stock CV passes the CI assertion.
Ten new test_verify_pdf.py cases. Proposed by 9scorp4 in Discussion #385.
* fix(verify_pdf): --ascii-dates only joins a year and a dash across horizontal whitespace
`\s*` between the year and the dash also matched a newline, so a heading
ending in an ASCII date ("Heading 1988-1994\n- note") joined the dash
that merely opened the next line and was reported as a U+2013 range; the
mirror shape ("Six Sigma -\n2016 onwards") did the same. Both are real
text-layer shapes: a dash used as a bullet, a wrapped prose line.
Match horizontal whitespace only ([^\S\n]*). Every existing case still
passes; three fixtures pin the two newline shapes and the same-line
spaced dash that must still be caught.
Found in review by yang2632, with the regex and the fixture.
normalizeId took the first 6+-digit path segment from ANY URL, so a
Greenhouse or Lever apply link - the kind a posting's own page hands out
and a user pastes back into detail - was reduced to that number and the
handler fetched jobPosting/<number> from LinkedIn: whatever job carried
that id came back with exit 0, or NOT_FOUND if none did, never an error
about the input. Driving the real handler with a stubbed fetch,
https://boards.greenhouse.io/acme/jobs/4567890 requested
.../jobPosting/4567890. Every other portal CLI rejects an off-host detail
URL with BAD_ID (the #447 shape); linkedin was the one still trusting the
digits.
Parse URLs for real: a linkedin.com host (apex or any subdomain, scheme
optional) plus a /jobs/view/<slug->id path yields the id; other hosts,
look-alike and userinfo hosts, and linkedin.com profile or search URLs
exit 1 with the stderr-JSON BAD_ID contract before any request. Bare
ids, URNs, and slash-free title slugs are unchanged.
Four new normalizeId cases and a new detail-input.test.ts that drives
runDetail and the CLI with fetch stubbed; five cases fail on the old
pattern.
`tools/verify_layout.py`'s module docstring explains which documents trigger
the Poppler `-bbox` crash, and names the macro whose absence causes it:
`\hypersetup{pdftitle=...}`. In a non-raw docstring `\h` is an invalid escape
sequence, so compiling the module prints
SyntaxWarning: "\h" is an invalid escape sequence.
Such sequences will not work in the future. Did you mean "\\h"?
on Python 3.12+ (CPython gh-98401 promoted the long-standing
DeprecationWarning to a SyntaxWarning). The 3.15 language reference still
documents the sequence as a SyntaxWarning, with a SyntaxError only "in a
future Python version".
This is on the real path, not a constructed one: /apply Step 5b runs
`python3 tools/verify_layout.py <pdf>` once per document, so the first
invocation after any source change prints the warning into the middle of the
layout report a user is reading to judge their CV. It is also visible in this
repo's own CI log, where `python -m unittest discover -s tests -t . -v` prints
it above the test results.
Reproduce on master:
python3 -W error::SyntaxWarning -c "compile(open('tools/verify_layout.py',
encoding='utf-8').read(), 'tools/verify_layout.py', 'exec')"
raises on master; after this change it is silent.
Escaping (rather than making the docstring raw) keeps the surrounding prose
unchanged, and the single-backtick markup matches the other inline literals in
the same docstring. The rendered text still reads `\hypersetup{...}`.
The new `ToolsCompileWithoutWarnings` case is a pure `compile()` of the source
text of every `tools/*.py` inside `warnings.catch_warnings(record=True)` and
asserts no `SyntaxWarning` or `DeprecationWarning` (3.10 and 3.11 report the
same escape as a DeprecationWarning) - no cache file, no temp file, so it runs
the same on Windows (py_compile's atomic rename over an open
NamedTemporaryFile is what the first version tripped on). It fails on master
and passes here on every Python in the CI matrix, and it guards the next
docstring that quotes a macro.
Verification on Python 3.10, 3.11, 3.12, 3.13 and 3.14: `python3 -m unittest discover -s
tests -t .` (488 tests OK; master is 487), `python3 -W error::SyntaxWarning`
compile of every `tools/*.py`, `tests/test_changelog_structure.py`.
A piped stdout on Windows defaults to the ANSI code page (cp1252 on most
Western installs). rank_state.py printed titles and companies with
ensure_ascii=False, so one Cyrillic, CJK, Devanagari, Polish or Turkish
posting ended /rank with UnicodeEncodeError. job_key --audit, salary_lookup,
convert_salary_excel and verify_layout failed the same way.
Each tool now reconfigures stdout and stderr to UTF-8 at entry. The new
tests run every tool in a child process with a cp1252 stdout forced through
PYTHONIOENCODING, so Linux CI reproduces the Windows failure.
Claude-Session: https://claude.ai/code/session_01WPXraLZ4i7UW4xGn9tVcox
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
* feat(add-template): keep a registered template's intermediates in build/
A template registered via /add-template declared a bare compile command, so
every new template scattered .aux/.log/.out beside its source, and Step 4 had to
enumerate six LaTeX extensions to clean up after its test compile.
- The elicited LaTeX command now redirects intermediates to build/ with the
portable -output-directory (not the MiKTeX-only -aux-directory), creates the
folder (TeX Live does not), and moves the PDF back beside the source where
/apply reads it. A leading rm -f ensures a failed compile, which skips the mv,
never leaves a stale PDF there.
- Toolchains with nothing to redirect (typst compile) keep their command, and
registration never blocks on the rule. If the redirect alone breaks a
template (e.g. \include from a subfolder), it is dropped and recorded under
Known pitfalls.
- The manifest example shows the pattern, Step 4 cleanup deletes build/
instead of listing extensions, and the ACTIVE-TEMPLATE block tells /apply to
run the command from the output directory and remove build/ in Step 5e.
- build/ is gitignored.
Stock templates are untouched, so /apply Step 5e stays.
* docs(add-template): drop the MiKTeX aside from the build directory rule
* test(add-template): pin the build directory compile command
Step 5b's prose said "Page count is not checked here - that is
verify_pdf.py --pages's job, and Step 5d already runs it", and
verify_layout.py's docstring declines to measure page count for the same
reason. Step 5d's only verify_pdf.py call is --dump-text, and no step in
the workflow passed --pages at all (only the upstream-only CI assertion on
the stock examples does), so the hard 2-page CV and 1-page cover letter
limits were enforced by nothing but the visual PDF read - the "measure
first, then look" failure 5b was written to stop.
5b now runs verify_pdf.py --pages 2 on the CV and --pages 1 on the cover
letter ahead of verify_layout.py, names the ACTIVE-TEMPLATE page limit as
the substitute for a custom template, and the deferral sentence points at
those lines instead of at 5d.
tests/test_apply_page_count.py pins the invocations, their counts, their
order relative to the layout measurement, and that no prose defers the
check to a step that does not run it; all four cases fail on master.
format_entry appended "* N/A = Too few employees to publish (privacy)"
under every category table, including one where every row has an index,
so the output asserted a suppression that never happened - the residual
noted on #470. Set a flag in the N/A* branch and print the footnote only
when it fired; a table with a suppressed row renders exactly as before.
Two FormatEntryTests cases pin both directions; the "omitted" case fails
on master.
The pairing loop required a non-empty derived category name on both
sides, but a header with no category word - "Count" + "Index", Danish
"Antal" + "Lønindeks" - strips to an empty name, so the simplest layout
the README advertises ("auto-pairs count/index columns") came out as two
unrelated standalone categories:
{"count": {"count": 500}, "index": {"index": 108.5}}
salary_lookup then rendered a "Count 500 N/A*" row above an
"Index - 108.5" row, and its footnote read the N/A* as "too few
employees to publish (privacy)" - a false statement about a company whose
headcount is in the file, shown during /apply's salary step. Any suffix
("Antal alle") made pairing work, which is why the shipped tests, all
suffixed, never saw it.
Pair on equal derived names, empty included, and give the nameless pair
the README's top-level category name (all_employees). A bare "Antal" with
no bare index column still stays a standalone count; named pairs
alongside are untouched.
Four new cases in test_convert_salary_excel.py, one of them rendering the
converter's output through salary_lookup.format_entry; all four fail
against the old pairing rule.
* fix(gmail-sync): strip CSV-breaking characters from the email subject
Step 7a interpolated the raw subject line of a received email into the
`notes` column of job_search_tracker.csv. No writer here emits a quoted
tracker field, so an unescaped comma splits the row - for csv.DictReader
just as much as for a naive split, which matters because
tools/rank_state.py is the repo's only machine reader and uses exactly
that. `notes` is column 10 of 14, so a subject as ordinary as
"Re: Your application, Data Scientist" shifted cv_file,
cover_letter_file and source a column left. A line break is worse: it
ends the row and starts a second one.
The rule now sits on the append instruction itself rather than in a
general note a writer can miss. The subject survives verbatim in the
archive's outcome.md, which is Markdown and carries no such constraint.
/outcome Step 4 (outcome.md:195) is also free-form and has the same
exposure, but its text is model-authored in a turn the user is watching
rather than copied from third-party mail unattended. Left out
deliberately, to be filed separately.
* fix(outcome): keep the Step 4 tracker note free of CSV-breaking characters
/outcome Step 4 appended "a short dated note" to `notes` with no
constraint on its content, the same exposure /gmail-sync Step 7a had:
nothing quotes a tracker field, so `rejected, no feedback given` shifts
cv_file, cover_letter_file and source a column left under
csv.DictReader, and a line break ends the row. The append instruction
now requires a note with no commas, double quotes or line breaks.
Folded in at the maintainer's request on #455 so one entry and one rule
cover both free-form writers. The CSV-safety tests move out of
test_gmail_sync_command.py into test_tracker_notes_csv_safe.py, where a
CASES table pins the rule on each writer's append line.
verify_layout.py's skipped: message blamed only the xpdf-based pdftotext that
Git for Windows puts ahead of Poppler in PATH. A real Poppler can abort too:
26.0x before 26.05 crashes -bbox/-bbox-layout/-htmlmeta on a PDF whose Info
dictionary carries an empty string in any field, and hyperref writes exactly
that for every field it does not set. A lualatex/pdflatex document built with
hyperref and no \hypersetup{pdftitle=...} - an ordinary /add-template CV
template, not a malformed one - hits this with a working Poppler installed,
and the old message sent the reader to check their PATH when nothing was
wrong with it.
Named both causes in the raised message, the code comment above it, and the
module docstring. Behavior is unchanged: either cause still degrades to
skipped: exit 2, not exit 1, since a broken extractor is still not a broken
document.
Added test_poppler_abort_on_empty_info_string_names_that_cause_too, mocking
subprocess.run to raise CalledProcessError the way a real Poppler 26.0x abort
does (exit 1, a libc++abi out_of_range trace on stderr) rather than the
xpdf case's exit 99 - the two failures are the same exception type with
different exit codes and stderr text, so the message has to actually
distinguish them, not just catch the one class. Negative control: this test
fails against the pre-fix message with "'Poppler aborted' not found in
'...a pdftotext without -bbox is usually the xpdf build...'".
Thanks to @main-sounds-audio for the report and the isolated repro (which
field, which Poppler modes, five runs of five) that pinned this to Poppler's
own upstream regression (issue #1699, fixed in 26.05.0) rather than an
extraction-library swap.
Fixes#451
TemplatesStillCarryThePlaceholders asserts 05-cv-templates.md and 06-cover-letter-templates.md still carry their [FIRST_NAME]/[YOUR_*] tokens. /setup's documented Step 3.5/3.6 personalisation replaces exactly those tokens, so python-tests failed permanently on any personalized fork. Same @unittest.skipIf on GITHUB_REPOSITORY (defaulting to upstream when unset) as test_placeholder_integrity.py received in #407; that fix landed three days before this guard, which did not pick up the pattern.
Co-authored-by: Nathan No-ot <244263078+Nnoot02@users.noreply.github.com>
`normalize_text()` folded whitespace only, so `--contains` compared what a
user types against what LaTeX renders. The stock CV compiled with the
documented lualatex command turns `'` into U+2019 and `--` into U+2013, so
`--contains "Master's degree"` and `--contains "2016-2024"` both reported
the keyword missing from a document that plainly contains it, through both
extractors. The documented remedy for a missing keyword is to add it, which
is the one thing the ATS section forbids.
Fold both sides at comparison time: NFC, then curly apostrophes and quotes
to ASCII, en/em dashes to `-`, no-break space to space. `--dump-text` still
writes the raw layer - that is what an ATS parses, and the date-range rule
in 05-cv-templates.md needs the raw en-dash visible there.
Separately, pdflatex without T1 font encoding stores accents decomposed
(`e` + U+0300). NFC repairs the pdftotext side of that, but pypdf reads the
same layer as `Z¨ urich` with a spacing accent, which no fold recovers.
moderncv 2.5 loads T1 itself under pdflatex; the apt-packaged 2.3.1 does
not - reproduced by compiling the template against moderncv v2.3.1 with
pdflatex (before: U+0308/U+0300 in pdftotext, `Z¨ urich` in pypdf; after:
U+00FC/U+00E8 in both). The template and the guide's preamble gain
`\ifpdftex\usepackage[T1]{fontenc}\fi`; the lualatex text layer is
byte-identical before and after.
Tests: ten new cases in test_verify_pdf.py (the fold-through and
normalize_text ones fail on the whitespace-only code) and a
test_latex_guidance.py guard that the fontenc line exists and stays inside
the pdflatex branch. framework_version 1.4.3 -> 1.4.4 on 05-cv-templates.md.
Reported and diagnosed by 9scorp4 in Discussions #385 and #384.
`detail` called fetch() directly rather than going through the CLI's own
request wrappers, so it had none of the three things apiFetch/apiPost
guarantee: no 429/5xx retry loop, a hand-inlined User-Agent that would
drift from the exported USER_AGENT, and a timeout no wrapper test covered.
A rate-limited detail page wrote API_ERROR and exited after ONE attempt;
jobnet, jobbank, jobindex, linkedin, and freehire all retry up to six
times on the same response. /scrape calls detail once per shortlisted
posting, so a burst that tripped jobdanmark's limiter dropped those
postings (no description, no deadline) while any other portal rode it out.
Demonstrated by driving the real command handler with a stubbed 429 and
instant timers: 1 fetch attempt and exit 1 before, 7 after (initial try
plus six retries, the contract's schedule).
Add htmlFetch to helpers.ts with the same backoff, timeout, and shared
User-Agent as the JSON wrappers - 404 returns null so detail keeps its
NOT_FOUND contract - and route detail through it. The retry-backoff,
user-agent, and request-timeout suites now cover all three wrappers, and
the new detail-backoff.test.ts exercises the handler path itself; its two
retry cases fail against the bare fetch().
detail fetched any http(s) input verbatim with no host check and, when
the path didn't match, silently used the whole input URL as the job id -
a non-posting page came back as a well-formed fake posting with exit 0.
buildUrl now requires a jobindex.dk host and a /jobannonce/<id> path,
rebuilds the fetch URL from the extracted id, and exits 1 with the
stderr-JSON BAD_ID contract otherwise; bare ids stay permissive
slash-free tokens per the jobnet precedent. Eight cases in the new
detail-input.test.ts; the five rejection/canonicalization cases fail
against the verbatim unguarded extraction.
When a tracker row's cv_file/cover_letter_file columns are empty, /outcome
and /interview fell back to a company-prefix glob (cv/main_<company>*.tex).
/apply names drafts main_<company>_<role><CV_EXT>, so two roles at one
company both match that glob. /outcome copied whichever the filesystem
returned first into the archive as cv_draft.tex - the file whose stated
purpose is to record what was actually submitted - and its own "leave an
existing archived file" rule then made the wrong copy permanent.
Both fallbacks now glob cv/main_<company>_<role>.* and
cover_letters/cover_<company>_<role>.*, deriving the stem by the Subfolder
naming rule in documents/README.md rather than restating it, and skip with
a note instead of widening the search. Dropping the hardcoded .tex also
makes a template registered by /add-template findable.
The dot before the extension wildcard matters: a bare trailing * also
absorbs a longer role, so ML Engineer and ML Engineer II at one company
would collide the same way the company-prefix glob did.
* feat(layout): measure compiled PDF layout instead of eyeballing it
/apply Step 5b asks for layout properties and executes none of them: they are
checked by reading the rendered page, which is exactly how they get missed.
The failure that motivates this is silent under every existing check. A moderncv
\cventry renders as a tabular, so an entry is one unbreakable block; when it does
not fit in the space left, the whole entry moves to the next page and leaves a
hole behind. The document still compiles, still reports the expected page count,
and still passes tools/verify_pdf.py. Observed in the wild at 273pt, roughly 19
blank lines, mid-page, on a CV whose visual read looked fine.
tools/verify_layout.py reports per page where the text starts and stops, bottom
whitespace as a share of page height, and the largest gap between lines, then
exits 1 on a hole over 100pt, a non-final page ending more than 25% early, body
text colliding with the page-number footer, a final page more than 35% empty, or
an entry header or section heading stranded at a page break.
Page count is deliberately not checked here - verify_pdf.py --pages already does
that, and two implementations of one rule drift. Geometry comes from Poppler
pdftotext -bbox, already a dependency; a missing Poppler exits 2 with "skipped:"
rather than failing the run.
Tests build synthetic Page/Line geometry, so the suite needs neither Poppler nor
a LaTeX toolchain and runs on the existing 3.10-3.14 matrix.
* fix(layout): survive Windows encoding and a pdftotext without -bbox
Review found two failures on the repo's primary platform:
subprocess.run(..., text=True) decoded pdftotext's UTF-8 output with the
Windows ANSI codepage and crashed on the stock cv/main_example.pdf. It now
passes encoding="utf-8" with errors="replace" at the call site, matching the
fix verify_pdf.py already carries from #369.
Git for Windows ships an xpdf-based pdftotext that shadows Poppler in a
default PATH and has no -bbox flag; it exited 99, the CalledProcessError
escaped, and the run ended in exit 1 - indistinguishable from a real layout
problem, which would send /apply chasing a phantom hole. That now routes to
the existing "skipped:" exit 2 path with a message naming the likely cause,
covered by three tests on the extractor-failure path.
Also from review:
- .claude/settings.json and security_guards.py gain the verify_layout.py
permission entries. apply.md Step 5b runs the tool on every /apply, so
without them every run prompts.
- The docstring and CHANGELOG no longer call Poppler a dependency
verify_pdf.py relies on. Since #369 verify_pdf prefers pypdf and Poppler is
the fallback; word bboxes have no pypdf equivalent, so this is the one step
that still wants it, and that is now what the text says.
- Docstring and apply.md state that the thresholds are calibrated for the
stock moderncv and cover.cls geometry, and that the shipped example CV
fails the thin-final-page rule by design.
- largest_gap documents that it measures top-to-top, so a tall line inflates
the gap by its own height - over-detection, the safe direction.
- The test module imports via tools.verify_layout like test_verify_pdf.py
instead of sys.path.insert.
* fix(verify_layout): quote only the first stderr line in the skip message (#378)
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013fqqLgQSnwgWkv98twQhHi
---------
Co-authored-by: Mads Lorentzen <madslorentzen_17@hotmail.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
* fix(scraper): make seen_jobs.json keys a pure function of the posting, and clean up the drift
The dedup key was prose only ("<url_or_company_title_key>"), so different
/scrape runs slugified company+title differently and the state file
accumulated two failures: keys carrying "&", "/", "," and ":" that break
the archive-folder path /apply and /outcome derive from company+role
(documents/README.md's subfolder rule exists because of exactly this),
and the same posting stored twice under two different truncations of a
long title (two Deloitte entries, one job, one URL).
tools/job_key.py makes the key a pure, deterministic function of
company+title+url: a strict allowlist slug, length-capped with a hash of
the full slug so truncation never collides across runs, and a fallback
to the portal's numeric job id when a non-Latin title slugifies to
nothing (a real prior entry, "securion_", would have collided with
every future non-Latin posting from that company).
--audit finds both failure classes in an existing seen_jobs.json without
guessing at a fix: malformed keys (real damage), a legacy three-part
company_title_location shape (harmless but not what the current rule
produces, so it silently re-duplicates on the next scrape), and
duplicate URLs. Ran it against this workspace's file and re-keyed the 15
entries it found - 7 malformed, 8 legacy-shape - verified byte-for-byte
against the pre-cleanup copy that no entry's data changed, only its key.
tests/test_job_key.py (16 tests) covers the slugify rules, the
truncation-hash behavior, both non-Latin fallback paths, and the audit
CLI's exit codes.
* fix(scrape): call the key helper from Step 4 instead of slugifying ad hoc
The helper added in the previous commit is only load-bearing if the spec
calls it. Step 4 described the key as prose ("<url_or_company_title_key>"),
which is what let each run slugify its own way. Step 4 now names the
command, and the schema shows the key's provenance.
Renumbers the trailing list item; no other behaviour in the step changes.
* docs(changelog): record the job-key rule under Unreleased
* fix(scrape): preserve dedup continuity across key rule
* changelog: note that existing seen_jobs.json entries need no migration (#441)
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013fqqLgQSnwgWkv98twQhHi
---------
Co-authored-by: nox <nox@Mac.home>
Co-authored-by: Mads Lorentzen <madslorentzen_17@hotmail.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Both branches predate the v1.7.1 cut, so their hunks' context pointed at the
old [Unreleased] "### Added" header, which the cut had rolled into [1.7.1].
The rebases applied cleanly and landed the entries inside the release. They
belong under [Unreleased].
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013fqqLgQSnwgWkv98twQhHi
* feat(outcome): add stale sweep branch for batch-resolving quiet applications
- Add /outcome stale [N] and /outcome sweep [N] to Step 0
- Offer stale sweep in Step 1.3 when rows are quiet 60+ days
- Add Step 2c Stale Sweep Branch with interactive all/select/skip confirmation
- Batch-resolve qualifying open applications to canonical no_response
- Update tracker status and append dated notes; update archive outcome.md
- Add Rule 9 to Important Rules
- Add tests/test_outcome_stale.py and update CHANGELOG.md
* test(outcome): drop auxiliary candidate filter from spec test per review
* fix(web-research): fail loudly when $SCRATCHPAD is unset
The two runnable snippets in 09-web-research.md both start with
`cd "$SCRATCHPAD"`, but nothing in the repo ever sets that variable.
With it unset the command expands to `cd ""`, which succeeds and leaves
the shell in the current directory, so `page.html` and the extracted
text land wherever the command was run from. In practice that is the
repo checkout, which is exactly what the paragraph directly beneath the
curl block forbids: "Write to the session scratchpad directory, never
into the repo."
Guarding with `${SCRATCHPAD:?...}` turns a silent write into the repo
into an immediate, self-explaining failure. The message names where the
value comes from so the reader can set it and re-run.
* docs(changelog): record the $SCRATCHPAD guard under Unreleased
---------
Co-authored-by: nox <nox@Mac.home>
Rolls [Unreleased] into [1.7.1] - 2026-09-06 and moves the compare links.
Keeps an empty [Unreleased] heading above the release, per Keep a Changelog,
and teaches the CHANGELOG structure guard that an absent [Unreleased]
section is nothing to check rather than an error.
Claude-Session: https://claude.ai/code/session_013fqqLgQSnwgWkv98twQhHi
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
/setup Step 3 personalised cv/main_example.tex but never the LaTeX contact
blocks embedded in 05-cv-templates.md and 06-cover-letter-templates.md, the
two files /apply actually compiles from; 06 was not a Step 3 target at all.
Step 3.5 now names the 05 contact tokens, a new Step 3.6 covers the 06
contact line and signature, the completion summary lists 06, and /reset
restores both blocks instead of listing 06 as framework-only (the existing
/reset coverage test forced that half).
tests/test_changelog_structure.py checks [Unreleased] on every PR for
duplicate headings, unknown headings, orphan entries and conflict markers -
the #425 duplicate-heading shape that was fixed by hand at merge time.
Claude-Session: https://claude.ai/code/session_013fqqLgQSnwgWkv98twQhHi
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
* fix(rank): move seen_jobs.json read/write off the state file's own critical path (#395)
/rank's Step 1 read the whole of seen_jobs.json into the conversation to
select candidates by eye, and Step 4 emitted it back to record scores.
That cost is paid on every run regardless of how many jobs are scored,
and it grows for the life of the workspace, since the file is
append-only and most stored entries are `skipped`.
tools/rank_state.py moves that traffic into code:
- `candidates` selects entries per Step 1's existing rules (status
filter, tracker exclusion, focus filter, `--limit`/`--all` from #424)
and projects only the fields a scoring agent needs.
- `sweep` runs rule 6's expiry pass over entries the run did not
re-score - a stored-date comparison, no fetch, no agent - preserving
its defensive parsing of non-ISO deadlines and its `--all`
reversibility.
- `apply` writes scoring results back atomically and prints the
ranked/vetoed/expired rows Step 5's report is built from, preserving
Step 4's existing write-back rules exactly: the `location` ->
`location_verdict` legacy migration, the deadline
null-is-not-a-correction rule, verbatim strengths/gaps persistence,
and idempotent re-scoring.
Step 1, Step 3's rule 6, and Step 4 now route through the tool instead
of describing a manual read/write. Nothing about scoring policy changes
- no new status, no new persisted field, no change to what counts as a
veto. The tracker stays read-only and every write is atomic (temp file
+ rename).
tests/test_rank_state.py (25 tests) covers the three subcommands
directly. The new spec-guard class in test_rank_command.py derives the
fields Step 4 must preserve from Step 2's own JSON schema block rather
than retyping them as a second list, so a future edit to that contract
is what the test reads instead of something that can drift from it.
* fix(rank): add CHANGELOG entry and remove the undefined $SCRATCHPAD reference
Two mechanical fixes from review:
- Step 4 named the results hand-off file via $SCRATCHPAD, a variable
nothing in the repo defines - a reader following the spec literally
has no path to substitute. Named the location in prose instead (a
temporary file outside the repo tree, never committed) and replaced
the shell-variable-looking path in the example command with an
explicit placeholder.
- Added the [Unreleased] entry this change was missing; the one
already in the diff belongs to #424.
* changelog: fold the #395 entry into the existing Fixed section
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013fqqLgQSnwgWkv98twQhHi
---------
Co-authored-by: nox <nox@Mac.home>
Co-authored-by: Mads Lorentzen <madslorentzen17@gmail.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
The guard in the four bunli-based CLIs inspected only tokens starting
with `--`, so an undefined short flag bypassed it: bunli discarded it,
the search ran unfiltered, and the CLI exited 0. Live against jobnet,
`search -q "sygeplejerske"` returned all 18,179 ads as a successful
search against 667 for the real `--search-string` query - the same shape
as review finding F13 that motivated the guard.
Both dash forms are now checked. Declared shorts (jobindex's -q) and
bunli's built-in -h/-v stay valid. A negative number is rejected too:
bunli discards a `-`-prefixed token rather than consuming it as the
previous flag's value, so `--radius -5` silently fell back to the
default instead of failing its own min(1) schema; a value that must
begin with a dash uses the `--flag=value` form.
Fixes#426.
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
The filter derefed item.text.toLowerCase() from a cast API response on
the same line that already guards g.items ?? [], so one item with a null
or missing text threw TypeError and the whole command exited 1 as
API_ERROR. The filter is extracted into an exported
filterAutocompleteGroups (the jobnet testability pattern), text is typed
nullable so the compiler enforces the guard, and an item without usable
text is skipped: it can never match the required non-empty query, so
downstream output never sees one. The null-text case was verified to
fail against the verbatim unguarded extraction with the production
TypeError. Closes out the #416/#418 audit.
date: job.publicationDate.slice(0, 10) trusted a TypeScript interface
claim nothing validates at runtime: apiFetch casts the JSON body, so one
ad with a null publication date threw TypeError inside the jobAds map
and the whole search exited 1 as API_ERROR. The neighboring
applicationDeadline field was already null-guarded with a 1900-01-01
sentinel. publicationDate is now typed nullable so the compiler enforces
the guard, and the ad degrades per-item to date: null. New test verified
to fail on the unfixed code with the exact production TypeError.
Add @unittest.skipIf on GITHUB_REPOSITORY to TestCvSentinelsAreDataLocated
and TestProfileSentinelIsDataLocated so python-tests matches the upstream-only
placeholder-integrity job. Default unset GITHUB_REPOSITORY to upstream so local
pristine-template runs still execute the guards.
FixesMadsLorentzen/ai-job-search#405
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: shahidbeig-a11y <shahidbeig-a11y@users.noreply.github.com>
* fix(salary): require corroboration before accepting a header row
Header-row detection accepted the first row (of the first 10) where any
cell merely contained a company-pattern word - no check that the row
actually looked like a header. A source-citation row above the real
header table (standard in real Danish union/statistics exports, e.g.
"Kilde: ... opdelt efter arbejdsgiver ...") tripped it purely because
"arbejdsgiver" appeared in prose. The real header row then parsed as
data (its "Firma" cell became a bogus company), and every genuine
company silently lost all its salary data - exit 0, no warning.
A candidate row is now only accepted when a second cell also matches a
city/count/index pattern, and a sheet that ends up with zero detected
salary columns prints a warning instead of reporting success silently.
Fixes#414.
* fix(salary): require cross-cell corroboration, fall back for untyped columns
Two edge cases found in review of the corroboration fix:
- Same-cell corroboration wasn't enough: a citation sentence can pack a
count-pattern word into the same sentence as the company-pattern one
("...opdelt efter arbejdsgiver, antal svar 1234"), which still passed
the gate. Corroboration must now come from a different cell.
- The corroboration requirement itself broke sheets whose only real
header has purely untyped salary columns (e.g. "Base pay 2025" /
"Bonus 2025" - neither matches a known city/count/index pattern), so
header detection found nothing at all. Falls back to the original
any-cell-mentions-company rule when the strict pass finds no row in
the first 10.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
new Date(<unparseable>) yields an Invalid Date whose toISOString()
throws RangeError, and normalizeSearchItem runs inside an unguarded
items.map(), so one malformed RSS item killed the entire search with
{"error": "Invalid Date", "code": "API_ERROR"} and exit 1. The
un-CDATA'd fallback capture in parseRssItems can deliver exactly such a
value. An unparseable pubDate now degrades to the same shape as an
absent one (posted "", date null); every other item survives. Three
new cases pin the malformed shapes, each failing on the unfixed code.
