Per Task 3 of the 2026-09-14 ASCII art upgrade plan.
- New module: src/main/AsciiArt.templates.js exporting ASCII_TEMPLATES
(19 entries) and getTemplate(name) accessor (returns '' for unknown).
- New test: tests/main/ascii-art.templates.test.js with 19 per-template
snapshots + unknown-name + keys-match assertions (21 tests total).
- 17 templates verbatim from src/ascii-generator.html:596-626
(arrow-right, arrow-down, decision, process, flowchart, sequence,
network, hierarchy, header, note, warning, info, divider, separator,
banner, checklist + the brief's own TEMPLATE_NAMES order).
- 2 templates verbatim from src/renderer.js:6542-6697
(progress-bar, table-simple).
- 1 newly authored: arrow-up (completes the arrow triplet; does not
exist in either source).
Amit Haridas
Widen height bound to 3-12 so the spec-named figlet fonts fit:
- Isometric1, Isometric2, Isometric3, Isometric4 (h=11)
- Calvin S (h=3)
Previously these were substituted with height-compliant alternatives
(Small Isometric1, Banner3-D, Henry 3D, Small Poison, Modular). Restoring
the spec-named fonts preserves the product intent: 4 distinct isometric
projections and the calvin-and-hobbes-style Calvin S font.
Amit Haridas
Replace the page-reload IPC handling in the theme-changed listener with an
in-place toggle of the preloaded <link id="theme-*"> tags. The helper scans
for a matching link first and only mutates disabled states when one exists,
so an unknown id (or main-process miss) short-circuits and preserves the
currently active theme. body.className is always set to the requested id so
legacy selectors keep matching.
Add tests/theme-renderer-apply.test.js with three cases covering: switching
to a fresh theme, idempotent re-switch, and the unknown-id short-circuit.
The test's local helper mirrors the renderer implementation so future drift
is caught at test time.
Amit Haridas
Preload one disabled <link id="theme-<id>"> per registered theme
between the structural CSS block and the KaTeX link. The renderer
will toggle the disabled attribute on the active link when the theme
changes. The 12 new-theme CSS files (catppuccin-*, one-light,
tokyo-night-storm, synthwave-84, outrun, winter-is-coming-*, solarized-dark-hc,
spring-light) are created in the next task; until then those 12 links
are inert (disabled) and 404 is harmless.
Amit Haridas
Wires the three registry-layer modules into src/main.js:
- Require ThemeRegistry + buildThemeMenu + the bootstrap side-effect
module at the top of the file so the registry is populated before
the menu is built.
- Replace the 109-line hardcoded View → Theme submenu block with a
buildThemeMenu({ setTheme, getCurrentThemeId }) call — themes are
now driven entirely from ThemeRegistry.list() / categories().
- Harden setTheme() to validate the incoming id against the registry
and fall back to 'atomonelight' if the stored id no longer exists
(e.g. after a downgrade or theme rename), with a console.warn so the
fallback is observable.
Pure wiring — no new behaviour, no unrelated edits. All 79 test suites
/ 931 tests remain green; lint and prettier checks clean.
Amit Haridas
Adds src/main/themeMenuBuilder.js — a pure module that consumes
ThemeRegistry.list() + ThemeRegistry.categories() and the injected
setTheme/getCurrentThemeId callbacks to produce the View → Theme
submenu's MenuItemTemplate[] in the same shape as the previous
hardcoded block in src/main.js:1137-1245.
- Radio-style items with checked=true on the active theme id
- Grouped by category in registry order with separators between
non-empty categories
- No Electron / electron-store imports — keeps the module pure and
unit-testable under @jest-environment node
- Tests use jest.resetModules() + per-test require to avoid the
module-cache leakage the bootstrap test surfaced in T2
Amit Haridas
Registers all 37 editor themes at startup via ThemeRegistry.bootstrap.js.
Side-effect module: requiring it populates the registry from a static
THEMES array (25 existing menu themes refactored into the registry +
12 new: Catppuccin x4, one-light, tokyo-night-storm, synthwave-84,
outrun, winter-is-coming x2, solarized-dark-hc, spring-light).
Categories split: 13 light / 22 dark / 1 high-contrast / 1 seasonal
(spring-light per spec).
Snapshot test asserts exact id order, shape validity, and category
counts. Tests use jest.resetModules() + per-test requires so the
bootstrap module re-evaluates its registration loop each run.
Amit Haridas
- 8 exports: register/unregister/list/get/categories/lightThemes/darkThemes/clear
- kebab-case id validator; category whitelist (light/dark/high-contrast/seasonal)
- duplicate-id and shape errors with descriptive messages
- pure CommonJS, no IO, no Electron deps — T2 bootstrap will register all 37 themes
- 10 Jest tests covering register/unregister/get/categories/light+dark filters
- full suite: 77 suites, 924 tests passing; lint + Prettier clean
Amit Haridas
Two step-by-step TDD plans derived from the 2026-09-14 design specs:
* theme-registry: 9 tasks, 70+ steps. Creates src/main/ThemeRegistry.js
(pure module), bootstrap with 37 themes, buildThemeMenu for main.js,
migrates 25 existing per-theme CSS blocks into src/styles/themes/<id>.css,
rewrites the renderer apply-theme function to toggle <link disabled>,
adds 12 new theme CSS files using shared token vocabulary.
* flowchart-editor: 8 tasks, 41 TDD steps. Pure renderer-side feature
(no main-process modules). Pure data store with IO injection, 5 SVG
shape functions, Mermaid translator, hand-rolled SVG canvas with
pointer events, sidebar panel wiring with debounced preview (250ms)
+ debounced persistence (500ms) + panel-scoped keyboard shortcuts.
Three minimal userData-path-validated IPC channels added for
persistence (renderer can't reach <userData> under the current
nodeIntegration:true security model without them).
Amit Haridas
Three new specs for the v4.8.0+ feature wave:
* Theme registry: replace hardcoded 25-theme menu in main.js with a pure
ThemeRegistry module + per-theme CSS file convention. Add 12 new themes
(Catppuccin x4, One Light, Tokyo Night Storm, Synthwave '84, Outrun,
Winter is Coming Light+Dark, Solarized Dark HC, Spring Light).
* ASCII art upgrade: consolidate dual implementations (standalone window
vs dead in-app modal) into a single path; add 12 hand-coded fonts +
figlet npm library for 400+ fonts; add copy/save/insert output
destinations; comprehensive tests for the previously-zero-coverage
textToASCII/createASCIIBox/getASCIITemplate machinery.
* Flow chart editor: sidebar panel with hand-rolled SVG canvas, node-graph
data model, drag/drop editing, live Mermaid source preview, undo/redo,
session persistence. Emits Mermaid which the existing preview pane
already renders natively.
Amit Haridas
Two more features from the deferred menu:
Daily-note template gallery:
- src/main/DailyNotesTemplates.js — pure module: listTemplates() /
saveTemplate() / deleteTemplate() / labelFor() with injectable IO.
- src/main/DailyNotes.js — openOrCreate() now accepts seedContent so a
non-default template can seed a NEW note (existing notes never get
clobbered).
- src/main.js — IPC channels daily-templates:list / save / delete /
apply. apply renders the chosen template (with {date}/{weekday}
substitution) and pipes through DailyNotes.openOrCreate.
- src/sidebar/daily-templates-panel.js — gallery UI: list, +New
(prompt for name + content), Use (applies to today's note),
delete (refuses to remove the last template so the default survives).
- src/renderer.js — registers the panel.
- src/index.html — icon (already added).
Pluggable DocQA engine (semantic search hook):
- src/main/SemanticEngine.js — engine interface with defaultEngine() (TF-idF,
always available) and neuralEngine() (lazy @xenova/transformers,
falls back gracefully when the dep is missing). getEngine(name)
resolves either.
- src/main/DocQA.js — ask() is now async and accepts an engine arg.
TF-idF path unchanged; neural path calls engine.rank(question, chunks)
directly. The chunk corpus is built up front regardless of engine so
ranking is consistent.
- src/main.js — doc-qa:ask IPC resolves the engine via SemanticEngine.getEngine(name)
before calling DocQA.ask. The renderer can pass {engine: 'transformers'}
to opt in once @xenova/transformers is installed.
Tests (51 new across this batch):
- tests/main/DailyNotesTemplates.test.js (18): labelFor separators /
edge cases / non-string safety, listTemplates empty / present / sort,
saveTemplate nested dir + .md extension + validation + null content,
deleteTemplate success / missing / validation.
- tests/daily-templates-panel.test.js (12): mount + empty state + list +
XSS safety, Use button (apply + error path), Delete button (success +
last-template guard), New template (save + cancel), refresh.
- tests/main/SemanticEngine.test.js (8): default engine shape + rank
matches WorkspaceSearch, getEngine for tf-idf / unknown / transformers
(graceful fallback when @xenova/transformers missing), parity check.
- DocQA: 5 new tests for engine arg (custom engine.rank called, default
fallback, neural hit shape translation); existing tests updated to
await the now-async ask().
Full suite: 76 suites, 914 tests, lint+format clean.
Activation for the neural engine:
npm install @xenova/transformers
(heavy; ~50 MiB with deps) — then 'transformers' is selectable in
doc-qa:ask. Until then, all calls use TF-idF transparently.
Amit Haridas
Daily notes panel:
- src/sidebar/daily-notes-panel.js — new sidebar panel that lists every
YYYY-MM-DD.md in the daily-notes dir (newest first), with a Today
button that creates/opens today's entry. Refresh button reloads.
Clicking a row calls onOpenFile(path). DOM is built with textContent
for dynamic fields — no XSS surface from a hostile filename.
- src/main.js — daily-notes:list IPC now returns absolute paths (joined
with the daily-notes dir) so the renderer can pass them straight to
open-file-path without re-synthesizing.
- src/renderer.js — registers the panel; icon for the daily-notes
button (calendar icon).
- src/index.html — calendar SVG icon for the daily-notes sidebar entry.
Q&A deep-link:
- src/main.js — open-file-path accepts either a string (legacy) or an
object {path, offset}. The offset is forwarded to the renderer via
file-opened.
- src/renderer.js — file-opened handler scrolls the editor to the
offset using EditorView.scrollIntoView(y: 'center') so the matching
passage lands in the middle of the visible area.
- search panel now passes {path, offset} to open-file-path so Q&A
results with chunk offsets jump straight to the passage.
Tests (12 new, tests/daily-notes-panel.test.js):
- mount, empty state, list rendering
- Today button → onOpenFile, 'Created today' / 'Today already exists' status
- error paths (openToday reject, listExisting reject, listExisting missing)
- click + Enter/Space on a row open the right path
- refresh() re-fetches, keeps status when keepStatus:true
- non-md entries filtered out
- XSS-safe textContent for dynamic data
Full suite: 73 suites, 873 tests, lint+format clean.
Amit Haridas
Same handler as the global Ctrl+Alt+D shortcut — opens (or creates)
the local-date YYYY-MM-DD.md and sends a file-opened IPC to the
renderer. Mirrors the Quick Note pattern in the same submenu (global
shortcut + menu entry, both pointing at openOrCreate).
Amit Haridas
- src/utils/csv-to-table.js — pure module. Auto-detects the delimiter
(tab wins over comma; comma wins over semicolon). RFC 4180-style
parsing handles quoted fields, escaped quotes (""), and CRLF. Cells
are escaped for markdown tables (\\ for backslashes, \| for pipes,
newlines stripped). Alignment: a column is right-aligned when every
non-empty cell matches a numeric pattern AND there are ≥2 rows or
≥1 multi-character cell (single-char numbers like "1" alone are
too ambiguous to call as numeric — could be labels). Single-row input
produces a header-only table (no fabricated "Column N" labels, no
body).
- src/editor/smart-paste.js — paste handler now tries CSV first (it
needs no async), falls back to the existing URL → title flow.
Tests (30 new, tests/csv-to-table.test.js):
- detectDelimiter: tab > comma > semicolon, empty/non-string safe
- parseRows: simple, RFC 4180 quoted, escaped quotes, CRLF, multi-line
- escapeCell: pipes, backslash-first escaping, newline strip, null/number
- csvToTable: simple CSV, single-row (header-only), TSV, empty input
- alignment: numeric detection (≥2 rows OR ≥1 multi-char cell), currency,
percentages, text columns stay left
- escaping inside cells (pipes, newlines)
- ragged-row padding
- looksLikeCsv: true for tab/comma multi-row, false for single row,
column-count mismatch, prose, very short input
Full suite: 72 suites, 861 tests, lint+format clean.
Amit Haridas
- src/main/UrlTitle.js — fetch a URL, return its <title>. Pure module
with injectable fetch for tests. Decodes named + numeric + hex entities
(AT&T, Café, —), strips the <title> tags, collapses
whitespace, caps the label at 200 chars. 5s timeout via AbortController,
2 MiB body cap to avoid OOM on big downloads, streaming reader with
overflow cancellation. Rejects non-http(s) URLs up front.
- src/main.js — IPC url-title:fetch proxies to fetchTitle.
- src/editor/smart-paste.js — CodeMirror 6 extension that detects
URL-only pastes (one URL, surrounded only by whitespace), inserts the
URL immediately, then async-rewrites the insertion range to
[Title](url) once the title arrives. Multi-line / prose pastes pass
through untouched.
- src/editor/codemirror-setup.js — accepts smartPasteFetcher option and
pushes the extension when provided.
- src/renderer.js — passes ipcRenderer.invoke('url-title:fetch') as
the fetcher.
- src/preload.js — url-title:fetch added to ALLOWED_SEND_CHANNELS.
- eslint.config.js — AbortController / TextDecoder / TextEncoder added
to globals (available in Node 20+ and Chromium).
Tests (34 new):
- tests/main/UrlTitle.test.js (24): isHttpUrl scheme filter, decodeTitle
named/numeric/hex entities + whitespace + non-string, extractTitleFromHtml
first match + case-insensitive + null fallback, fetchTitle success +
long-title cap + streaming body, all failure paths (non-http,
no-fetch, non-OK, wrong content-type, no <title>, network error,
body over maxBytes — including streaming overflow cancellation).
- tests/smart-paste.test.js (10): URL_ONLY_RE detection (bare URL,
path/query/fragment, whitespace padding, scheme rejection, embedded
rejection, empty/malformed), replacement format ([T](url), brackets
in title survive, query strings preserved).
Full suite: 71 suites, 831 tests, lint+format clean.
Amit Haridas
Hovering a footnote reference (rendered by marked-footnote as
<a data-footnote-ref href="#footnote-N">) now shows a small popover
with the footnote body text, matching the UX of Typora and Obsidian.
- src/renderer/footnote-preview.js — pure DOM module. Mounts a single
popover once per preview pane; mouseover delegates via Element.closest()
to the ref, lookups the matching <li id="footnote-N"> in the same pane,
strips the backref ↩, and positions above/below the cursor with edge
clamping. CSS.escape() fallback for non-browser environments.
- src/renderer.js — _renderPreview calls mountFootnotePreview once per
pane (gated by a dataset marker so re-renders don't accumulate
listeners).
- eslint.config.js — CSS + Element added to browser globals.
Tests (8 new, tests/footnote-preview.test.js): mount/unmount, delay +
timer behavior, mouseover/mouseout show/hide, dangling ref graceful
no-op, backref ↩ stripped from the displayed text, cancel-pending-show
when a new ref is hovered, idempotent remount guard.
Full suite: 69 suites, 797 tests, lint+format clean.
Amit Haridas
- src/utils/writing-stats.js — pure module: stripMarkdown() drops fenced
code blocks, inline code, image URLs, link URLs, headings, blockquote
markers, list bullets, and emphasis markers so the word count reflects
the prose a reader actually consumes (the convention used by Hemingway,
iA Writer). splitSentences / countSyllables use Flesch's standard
heuristics. computeStats returns wordCount, sentenceCount,
syllableCount, readingTimeMinutes (default 220 wpm),
fleschKincaidGrade, charCount.
- src/renderer.js — _updateStatusBar now calls computeStats and updates
three new status items: ~X min read, Grade N.N, and the existing Words
+ Chars counters use the stripped count.
- src/index.html — two new status items: #reading-time and #grade-level.
Tests (23 new, tests/writing-stats.test.js): markdown stripping (fenced,
inline, images, links, wikilinks, headings, lists, emphasis, HTML),
sentence splitter edge cases, syllable heuristic (short words, silent e,
empty), and computeStats with custom wpm + null-safe inputs.
Full suite: 68 suites, 789 tests, lint+format clean.
Amit Haridas
The workspace-search:query and doc-qa:ask IPC channels were reachable
from the renderer but had no UI. This commit wires them into a sidebar
panel that consumes both backends through a single input.
- src/sidebar/search-panel.js — one input, two modes:
Search (default): routes to workspace-search:query, returns file hits
Ask: routes to doc-qa:ask, returns chunk-level passages with offsets
Click a result → open the file (offset passed through for Q&A hits).
All dynamic content is escaped before innerHTML — hostile filenames
or snippets stay as text instead of becoming script/img nodes.
- src/renderer.js — registers the panel; reads the explorer's folder
input on each open so the search dir stays in sync.
- src/index.html — search sidebar icon (magnifier) next to the others.
Tests (18 new, tests/search-panel.test.js):
- mount + DOM structure
- Enter / click run → search() with query + dir
- result rendering (filePath, snippet, tag facet)
- onOpenFile receives (filePath, offset)
- Ask tab switches placeholder + routes to ask()
- Ask chunks carry +offset in the meta line
- empty query, no folder, search error → handled
- XSS: filename/snippet/tag with <script>, <img>, <unsafe> are escaped
- Escape clears, Clear button resets, host API (setDir/focus/clear)
Full suite: 67 suites, 766 tests, lint+format clean.
Amit Haridas
Three more features from the brainstorm menu, built on a shared search
algorithm so the codebase stays small.
- src/main/DailyNotes.js — Zettelkasten-style helper. One YYYY-MM-DD.md
per local date under <userData>/notes/daily/; loads skeleton from
<userData>/notes/templates/daily.md when present (built-in default
otherwise). openOrCreate never clobbers existing content.
- src/main/WorkspaceSearch.js — tag/wikilink-aware content search.
Pure module, injectable-IO tested. Query grammar: bare words,
#tag, @wikilink, "quoted phrases". Facets weight +3 each; prose
terms +1/occurrence capped at 5; edits within 7 days get a recency
nudge. Returns ranked results with snippets.
- src/main/DocQA.js — chunk-level Q&A wrapper over WorkspaceSearch.
cleanQuestion strips question words (what/how/why/...) and verb
noise (write/read/show/tell/...) so they don't drown the ranking.
Returns top-K passages instead of whole-file hits — multiple chunks
from the same file can appear in the answer.
- src/main.js — IPC: daily-notes:open-today, daily-notes:list,
workspace-search:query, doc-qa:ask. Path validation through the
existing validatePath gate; a global Ctrl+Alt+D shortcut creates
today's daily note from anywhere.
- src/preload.js — all four channels added to ALLOWED_SEND_CHANNELS.
Tests (56 new across the three modules):
- tests/main/DailyNotes.test.js (15): dateKey formatting, pathFor,
template load + {date}/{weekday} substitution, openOrCreate +
no-clobber, nested-dir creation, listExisting filtering,
isValidDir rejects NUL/non-string.
- tests/main/WorkspaceSearch.test.js (25): parseQuery grammar,
hasTag/hasWikilink word boundaries, scoreDocument scoring,
per-term spam cap, recency nudge, search ranking + limit +
empty-query short-circuit, bad-input safety.
- tests/main/DocQA.test.js (16): cleanQuestion stripping + facet
preservation, chunkDocument paragraph + hard-split, ask()
top-K, recency tiebreaker, missing-files fallback.
Full suite: 748 tests pass, 66 suites, lint+format clean.
Amit Haridas
VersionHistory snapshots the previous content on every explicit save — an
unsaved buffer is still lost on crash. AutosaveBuffer fills that gap.
- src/main/AutosaveBuffer.js — pure module mirroring VersionHistory's
injectable-IO pattern; one blob per doc path under
<userData>/autosave/by-path/<sha1>/recovery.md + meta.json. No history
(VersionHistory owns that) — just the latest dirty buffer.
- src/main.js — IPC channels autosave:write/read/clear/list; real paths
go through validatePath, synthetic 'untitled-tab-<id>' keys skip it.
- src/preload.js — added the four channels to ALLOWED_SEND_CHANNELS.
- src/renderer/autosave-client.js — debounced (2s) flush per tab +
periodic safety net (10s max age) + dirty-write retry on failure.
- src/renderer.js — register on tab create, unregister on close,
notifyChange piggybacks on performAutoSave's existing dirty-check,
clearForDocPath after a successful save, showAutosaveRecoveryBanner
on startup listing pending recoveries with Restore/Dismiss.
Tests (39 new):
- tests/main/AutosaveBuffer.test.js (19): round-trip, overwrite, isolation,
unicode/emoji, empty content, null coercion, ENOENT vs corrupt meta,
list ordering, corrupt-sibling skip, input validation, sha1 storage.
- tests/autosave-client.test.js (11): debounce, flushNow bypass,
no-path skip, clearForDocPath, list proxy, IPC error fallback,
unregister tear-down, failure-retry, periodic flush, idempotency.
Full suite: 692 tests pass, 63 suites, lint+format clean.
Amit Haridas
The renderer wires KaTeX via initMathSupport() and calls
window.renderMathInElement(...) inside _renderPreview() — there were no
tests pinning any of that contract. This commit adds tests/math-rendering.test.js:
- inline ($, escaped \() and display ($$, escaped \[) delimiters all
render valid TeX
- mixed prose + math preserves siblings (headings, lists, etc.)
- invalid LaTeX degrades gracefully under default throwOnError:false, so
the renderer's outer try/catch contract stays load-bearing
- dollar-heavy prose doesn't throw (the renderer's auto-render call must
be robust to anything in the doc)
- bundle wiring: katex.render() and renderMathInElement are exported,
katex.min.css ships in assets/, and index.html references it without a
CDN (the old jsdelivr link was removed in a prior hardening pass)
Amit Haridas
@ cantoo/pdf-lib 2.9.1 was already the dep in use; the encryption plumbing
and probe were already in place from the prior hardening pass. This commit:
- adds a test confirming executeOperation('permissions', ...) routes through
pdfSetPermissions and produces an encrypted PDF unlocked by the owner
password (mirrors the existing 'encrypt' route coverage)
- updates docs/superpowers/plans/2026-08-23-security-assessment-summary.md
to mark deferred risk D1 as resolved (the honest-failure message remains
as a fail-closed net for any future library regression)
35 PDF ops tests pass; lint/format clean.
Amit Haridas
Export themes:
- Six presets in the export dialog (basic + advanced modes) for PDF/DOCX:
Default, Modern, Classic, Sepia, Minimal, Elegant
- PDF: LaTeX header (xcolor/titlesec) recolors headings, adds section
rules and colored links — core-TeX packages only, hex-literal only
(no injection surface)
- DOCX: styles.xml surgery recolors Heading1-6/Title/Subtitle/Hyperlink
and swaps heading/body fonts; verified end-to-end against a real
pandoc-produced docx
- Themes ride along in export presets (unknown ids fall back to Default)
Windows CI fixes:
- pdfjs standardFontDataUrl now a file:// URL (backslash paths failed
pdfjs's trailing-slash validation, breaking extractText/extractImages)
- sharp temp cleanup EPERM retries; path-separator assertions; pdfjs
test timeouts raised; batch suite testTimeout 30s
648/648 tests green; 4.7.1 linux+win artifacts rebuilt.
pdfjs validates standardFontDataUrl as a URL ending in a forward slash —
our raw path with a trailing path.sep is invalid on Windows (C:\...\),
failing extractText/extractImages (and every test that verifies through
them) with 'Invalid factory url: must include trailing slash'. Linux and
macOS passed only because / is also a valid URL slash. Convert with
pathToFileURL() so every platform sends file:///.../standard_fonts/.
Also escape path.sep in PDFBatchOperations' sanitizer test regex — a bare
backslash made new RegExp() a syntax error on Windows.
The Windows runner's git checkout rewrote assets/fonts/FiraCode-LICENSE.txt
line endings (autocrlf), changing its bytes and tripping the SHA-256 pin in
download-tools.js. A .gitattributes marks the license (-text) and all font/
image binaries as never-normalized so every platform checks out identical
bytes.
Branding:
- All app icons regenerated from the new vector brand kit (M↓ mark):
icon.png/icon@2x (app + packaging), favicon.png, tray-icon.png, and the
full assets/icons/ size set — generate-icons.js now rasterizes
app-icon.svg directly (docico1.png removed)
- index.html gets proper favicon/apple-touch links from the kit
- Welcome tab hero shows the new mark; README gets the horizontal wordmark
- assets/logo.png (ConcreteInfo) intentionally untouched
CI release fixes (win/mac jobs were failing):
- FiraCode download moved from moving raw/master URLs (hash drifted
upstream, tripping the pin) to the immutable 6.2 release asset; repo
fonts updated to the pinned 6.2 bits
- macOS pandoc extractor locates the binary in the archive instead of
assuming a bin/ layout that the macOS zip doesn't have
Stray upload archive (markdown-converter-assets (1).zip) excluded.
- electron-builder config moves to electron-builder.config.js so markitdown
bundling is conditional per platform (PyInstaller only builds for the host
OS; a missing binary now logs a warning and ships without it instead of
failing the build); package.json static build section removed, all npm
scripts pointed at the config; third-party-licenses/ added to packaged files
- FIX (pre-existing): packaged apps looked for bundled pandoc in
resources/bin, but extraFiles land next to the executable (Contents/ on
macOS) — packaged builds silently fell back to system pandoc since 4.5.
New bundledToolDir() resolves the real location for pandoc + markitdown
- download-tools.js pins the win32 pandoc.exe SHA-256 (fetched + verified)
- sharp packaging test accepts sharp 0.35's versioned binding filename
- release.yml: bundle-markitdown step (best-effort) on every OS and a new
macOS job; release aggregates linux+windows+macos artifacts
Local release artifacts built and verified (dist/):
- MarkdownConverter-4.7.0.AppImage (363MB, pandoc+markitdown bundled,
packaged app boots clean, tools resolve at the fixed path)
- markdown-converter_4.7.0_amd64.deb (293MB)
- MarkdownConverter-Setup-4.7.0.exe (223MB), portable exe (223MB), zip
(305MB) — pandoc bundled; markitdown omitted (cannot cross-build),
legal docs verified inside app.asar
637/637 tests green; lint clean.
Bundle (v4.7.0):
- MarkItDown frozen to a single ~75MB per-platform binary (PyInstaller via
npm run bundle:markitdown; ML extras excluded) — built and verified
locally: HTML/XLSX/PDF conversions pass through the bundled binary, and
the app resolves bin/linux/markitdown first at runtime
- Packaging copies bundled markitdown alongside Pandoc for win/mac/linux;
FFmpeg/sharp/KaTeX/fonts were already bundled
Legal artifacts:
- THIRD-PARTY-NOTICES.md: complete license inventory of everything
distributed (binaries, npm runtime deps, fonts, embedded Python packages)
- SOURCES.md: GPL §3(b) written source offers for Pandoc 3.9.0.2,
ffmpeg-static's GPL build, and the PyInstaller bootloader; LGPL relinking
note for libvips
- third-party-licenses/: canonical GPL-2.0, LGPL-2.1, MPL-2.0, Apache-2.0,
OFL-1.1, PSF-Python texts
- Help > Third-Party Notices & Licenses: in-app viewer for both documents
- README: 'Bundled Dependencies, Legal Notices & Credits' section
Hardening:
- download-tools.js now SHA-256 pins every artifact, verifies after
download AND against the cache on every run, and hard-fails on mismatch
(closes security finding D6)
Large tools intentionally not bundled (documented): LibreOffice,
MiKTeX/TeX Live, ImageMagick, PlantUML+JRE, Calibre.
637/637 tests green; lint clean; clean boot; bundled binary verified.
- File → Import with MarkItDown (Any Format)…: PDF, DOCX, PPTX, XLSX,
Outlook .msg/.eml, EPUB, images, CSV/JSON/XML, ZIP (audio/OCR via the
[all] extras) — verified live against HTML, XLSX (our own exporter's
output), and PDF fixtures
- Command auto-resolution with caching: markitdown binary → python -m
markitdown → python3 -m markitdown
- SEC-1 argv discipline (execFile only, user paths never through a shell),
50MB cap, 120s timeout, sanitized errors that surface markitdown's own
"pip install 'markitdown[pdf]'" hints for missing format extras
- Output lands next to the source as <name>.md (numeric suffix, never
overwrites) and opens in a new tab; markitdown:available/convert IPC
allowlisted for renderer flows
- Help → Dependencies lists MarkItDown; README/UPDATES updated (v4.6.1)
12 new tests (629 green); lint clean; clean app boot
Anthropic-compatible provider:
- New 'anthropic-compatible' option for any base URL speaking the Anthropic
messages schema (LiteLLM proxies, Bedrock gateways, local servers)
- Sends x-api-key AND Bearer auth when a key is set (gateway-friendly,
harmless for the official API); keyless proxies supported
- Tolerates base URLs with or without a trailing /v1 segment
- Settings modal, manifest, and provider docs updated
Runtime bug fixes found by booting the app (run-to-verify pass):
- PDF editor: File > Open PDF sends operation=null which matched no switch
case and crashed on getElementById(undefined); now defaults to the merge
section
- backlinks-panel: wrong require depth (../../utils -> ../utils) threw at
panel registration time
- writing-studio stack was written against a synchronous settings backend but
the real one is IPC-backed: GoalTracker/SnapshotManager/ProjectManager and
all four panels now await; JSON.parse(Promise) crashes eliminated
- manuscript panel used window.prompt (unavailable in Electron); replaced
with an inline dialog
- collaboration comment-store/save-load made async to match its IPC IO
617/617 tests green; 4 consecutive clean app boots (no uncaught errors)
- AI Assistant plugin: multi-provider chat (OpenAI/Anthropic/Ollama/LM Studio),
summarize/improve/translate commands, proofread via ai:analyze; calls
proxied through main so API keys stay out of the renderer
- Collaboration plugin: anchor-based comments in .comments/ sidecars with
drift detection and F8 navigation
- Local knowledge base: [[wiki-links]] with click-to-create + Backlinks panel
- Crash recovery: debounced session snapshots with restore prompt on launch
- Version history: pre-save snapshots, History panel with restore/diff/delete
- Real PDF encryption: swap pdf-lib for @cantoo/pdf-lib (probe-driven UI)
- XLSX export (native workbooks via JSZip), ODT headers/footers + page size
- Offline KaTeX (bundled CSS+fonts), local-first PlantUML rendering
- Editor: vim mode toggle, snippet Tab-expansion, zen word-goal setter,
writing heatmap, writing-studio panels wired with rail icons
- Quick Note global scratchpad (Ctrl+Alt+Q), markdownconverter:// deep links,
REPL first-run confirmation
- Fix: Ctrl+Shift+P collision, pandoc converter availability check, CLI
dangling --css/--reference-doc flags, dead converter button
8 new test suites; 613 tests green; lint clean
build.asarUnpack only claimed node_modules/sharp/**, so electron-builder
pruned the @img/sharp-* optionalDependencies: the asar kept 34 pure-JS
@img entries while the bundled libvips shared libraries never shipped.
At boot sharp's loader fell back to a system-libvips-linked binding and
dlopen failed, crashing the main process. Claim the prebuilt packages
explicitly (@img/** and @napi-rs/**) per the sharp+electron-builder
recipe, and guard the built output with a packaging regression test.
Amit Haridas
A missing/pruned @img/sharp-* binding made the top-level require('sharp')
crash src/main.js at startup, killing the packaged app before any window.
Load sharp through a cached lazy getter instead; when the native module
cannot load, executeOperation resolves the honest failure shape
{ success: false, error: 'Image operations unavailable: <sanitized>' }
(free of absolute paths), mirroring PDFOperations' Task-27 precedent.
Amit Haridas
electron-builder's .snap (Squashfs) artifact in dist/ registered as an
obsolete Jest snapshot file, failing the suite exit code despite all
516 tests passing. modulePathIgnorePatterns keeps the snapshot scanner
out of build output.
Amit Haridas
Final-review nit: the hardcoded-list rationale covers the dialog path
only; the drop stands on the trusted-argv precedent for --convert-to.
Amit Haridas
The interval split mode looped for (i = 0; i < totalPages; i += interval),
which spins forever when interval <= 0. Both the single-file dialog and the
batch dialog can reach it (the batch dialog's validateOperationData only
checks truthiness, so -1 passes). Guard at the source in the main process:
reject non-positive or non-integer intervals before the loop, protecting
both paths and any future caller.
Amit Haridas
The get-pdf-form-fields IPC pair (handler in main.js, renderer invoke and
event.reply('pdf-form-fields')) was missing from ALLOWED_SEND_CHANNELS and
ALLOWED_RECEIVE_CHANNELS. Under the planned preload migration an unlisted
channel is silently blocked, so the form-field fill/flatten feature would
break once the main window stops using the inline shim. Placed adjacent to
the sibling get-pdf-page-count/pdf-page-count pair it was modeled on.
Amit Haridas
amitwh
Amit Haridas