Updates: 758 Commits | The Liberty Archive

September 9, 2026

Stop the filter bar flickering between its two heights
Outreach nightly log: 2026-09-09, ninth night all channels exhausted/capped/gated
Dead-link reclaim run 2026-09-09: mises.org blocked all 15 probes
Log 2026-09-09 forum links refresh
Refresh forum thread cards
End a chapter at the next chapter's anchor, not at a document boundary
Read a two-column PDF down its columns, not across the gutter

September 8, 2026

simplify permissions text
Record Reisman's 2025 signed licence, which does not commence yet
Sync the Government Against the Economy EPUB byte count
Convert Capitalism to EPUB and publish its reading pages
Redraw the Capitalism cover in the author's own palette
Document the scaled conversion timeout in the pdf-to-epub spec
Redirect the Capitalism cover brief to the 1996 jacket's palette
Write an original Reisman bio, and document the guide's new edition fields
Verify the edition of Reisman's Capitalism, and give the page a guide
Publish the permission basis in licenses.json and the runbooks
Document rights-holder permissions in the governing specs
Record rights-holder permissions, and Reisman's for Capitalism
Log 2026-09-08 outreach run: all channels still exhausted/capped/gated
Log 2026-09-08 forum links refresh
Refresh forum thread cards

September 7, 2026

Costs of War: replace the ratchet cover at David's request
Install 42 covers from batch 2, reject 7, and queue the Costs of War redraw
Cover proposals: batch 2 landed complete, 100 of 100
Install interim cover and refresh cache mtime
Hoppe: swap A Theory of Socialism and Capitalism to the other variant
Man, Economy, and Liberty: record the shade rejection as a dated note
Man, Economy, and Liberty: bar the lamp shade, brief the bare socket
Deploy: restore the SEO gate; make pre-push's skip explicit
Weekly KPI: /read/ growth continues, off-page links flat 8/111 two weeks running
commission_covers: --skip-proposed kept skipping the rejected books
Outreach drafts nightly: 2026-09-07, staged 0 (all channels exhausted/capped/gated)
Refresh forum thread cards

September 6, 2026

Redo Land Bubble cover: drop the gavel, bar it in the brief
Install 46 commissioned covers, reject 4, from the 100-image batch
Cover briefs: an impossible object is a tell, exclusion 7
Deploy: drop the WordOps origin assertion, keep the IIS one
Cover proposals: the 100-image batch landed complete
Record actual push outcome for tonight's outreach log entry
Outreach nightly log: 2026-09-06, sixth night all channels exhausted/capped/gated
Series rail: keep its heading at the width where the disclosure is forced open
Course page: the play control sits with its lecture, and progress counts a part-heard one
Lecture pages: series rail, player toolbar, share and auto-continue
Player dock: place the close button on the phone layout, not below it
Refresh forum thread cards

September 5, 2026

Course pages: shared player module, dock, progress and .m3u playlists
Publish wiki-link coverage to the dashboard, and spec the author rule
Link the archive and the quote site to each other on author pages
Match authors to the wiki by reading its index, not by guessing names
stats: icons for distribution cards and the two bare section headers
Spec the course and lecture page redesign: shared player, progress, dock, rail
commission_covers: codex reports two shapes of reset time, not one
Clear 249 catalogue descriptions that were PDF metadata, not prose
commission_covers: --wait-for-quota, so a large batch drains itself
commission_covers: a spent quota ended the whole run, silently
Add freecapitalists-cover-proposals-nightly, a draft-tier routine
commission_covers: make a batch resumable, because the quota ends it
Commission twelve covers from the generated brief corpus, ranks 12-28
top_books: report the commissioned tier it was blind to
outreach-status: date bump from the nightly routine
Cover briefs: final pass, and the last agent's skip logic held
Second pass over 22 briefs from the overlapping batch
Every book in the catalogue now has a cover brief: 969 of 969
Cover briefs: batch 026 of the generated corpus
Normalise all 46 Freeman volumes onto one register, and land batches 020, 023
Cover briefs: batch 024 of the generated corpus
Cover briefs: batch 025 of the generated corpus
Cover briefs: batch 022 of the generated corpus
Cover briefs: batch 019 of the generated corpus
Cover briefs: batch 028, the tail of the catalogue
Cover briefs: a numbered run of one publication is the exception to variety
Cover briefs: batch 015 of the generated corpus
Cover briefs: batches 016, 017 and 018 of the generated corpus
Cover briefs: batches 012 and 014 of the generated corpus
Add scripts/catalog_description_audit.py
Cover briefs: batch 013 of the generated corpus
Fix five em dashes, and stop demanding Context of briefs that cannot have one
Cover briefs: more of the generated corpus
Stop parsing the brief directories' own documentation as briefs
Commit the corpus runbook and the writer prompt it uses
Cover briefs: further batches of the generated corpus
Install a new America's Great Depression cover, and record a backend failure
Retire the second Introduction to Economics scan, keep the measured one
cover_check: the blind spot is bigger than one proposal
Cover briefs: make the corpus resumable, and land what survived the interruption
Fix five catalogue errors the cover-brief pass found, and credit co-authors
Cover briefs: batches 001 and 003 of the generated corpus
Install new covers for Human Action, Socialism, and Hoppe
Add scripts/cover_brief_audit.py, and fix two briefs it caught
cover_check: note the blind spot on typographic and flat-poster covers
Rewrite the three flagship briefs, and let codex generations overlap
commission-cover: judge against the shelf, and a corpus builder for briefs
Cover briefs: name a register, track spent motifs, exclude the generated look
Add scripts/cover_check.py: measure a cover at card size
Outreach nightly log: 2026-09-05, fifth night all channels exhausted/capped/gated
Refresh forum thread cards

September 4, 2026

Outreach nightly log: 2026-09-04, fourth night all channels exhausted/capped/gated
Refresh forum thread cards

September 3, 2026

Add skill runbooks for core content workflows
AI visibility test: first run, 2026-09-03
Record actual push outcome for tonight's outreach log entry
Outreach nightly log: 2026-09-03, third night all channels exhausted/capped/gated
Log 2026-09-03 forum-links refresh
Refresh forum thread cards

September 2, 2026

Outreach nightly log: 2026-09-02, all channels capped/exhausted/gated
Dead-link reclamation: run 2, ranks 17-31, 3 dead / 0 candidates
Refresh forum thread cards

September 1, 2026

fix grammar
Clean overly verbose copy
Revert "Add Top content section to /stats/: real Search Console clicks, above the method note"
Add Top content section to /stats/: real Search Console clicks, above the method note
Record actual push outcome for tonight's outreach log entry
Outreach nightly: log a genuine zero-item night, all channels exhausted/capped/gated
Log forum-links nightly refresh: 2026-09-01
Refresh forum thread cards

August 31, 2026

Weekly KPI: /read/ pages nearly double reach, author-page rebuild still undeployed
Outreach nightly: stage OpenDOAR, OER Commons, DOAB directory items
Log nightly forum-links refresh
Refresh forum thread cards

August 30, 2026

Outreach nightly: exhaust open-library queue, resolve OBP directory mechanism
Log forum thread cards refresh 2026-08-30
Refresh forum thread cards

August 29, 2026

Fix academic outreach channel: real drafts blocked by dead-end search
Record actual push outcome for tonight's outreach log entry
Outreach nightly: stage 3 Open Library work-record edits
Add competitor feature scan: full-text search is the top gap
Size the lecture-speaker author-page gap left open in today's entry
Record the 2026-08-29 two-engine pull and route the IndexNow runbook
Fix two byline defects in lecture and book titles
Add IndexNow submission, aimed at the engine that actually ranks this archive

August 28, 2026

Refresh forum thread cards

August 27, 2026

Point /stats/ at the companion wiki's own Special:Statistics
Wire publish_dashboard_stats.py into the nightly forum-links runbook
Update Open Library outreach status records
Record actual push outcome for tonight's outreach log entry
Outreach nightly: stage 3 more Open Library work-record edits
Refresh forum thread cards

August 26, 2026

Refresh outreach status timestamp
Outreach nightly: stage 3 Open Library work-record edits
Dead-link reclamation: first run, cursor at 15, zero candidates
Refresh forum thread cards

August 25, 2026

Match the catalog against a second forum's topics, for objectivism-online
Publish book/course-forum discussion-card coverage to dashboard.davidveksler.com
Update outreach-status.json
Refresh forum thread cards

August 24, 2026

Draft two institutional rights-holder emails, the ask that unlocks the most held text
Write up the 08-24 SEO read: where the full-text blockers actually are
Publish full text for The Progressive Era and two more; audit the shared-EPUB trap first
Make author pages read as bibliographies, and fix their inverted authorship schema
Record actual push outcome for tonight's outreach log entry
Stage nightly outreach: DataHoarder post and two Open Library edits
Refresh forum thread cards

August 23, 2026

Add sixth pipeline card: local forum AI on desktop GPU
Refresh forum thread cards

August 22, 2026

Update forum-activity.md with what actually shipped on the forum side
Add a live forum activity widget to the homepage
Widen the crosslink card to all three link directions
Create settings.local.json
Add forum-wiki crosslink release data
Add forum crosslinks pipeline to built page
Add a Companion forum section to llms.txt, next to the wiki pointer
Put a start-a-thread CTA on every book and course page, not just matched ones
Record that the first forum-links build went live
Correct the per-run topic-fetch figure to the measured 105
Link book and course pages to the forum threads that name them
Outreach nightly: second run this day, cap already spent, no new items
Fix what the outreach routine's first live run exposed
Stage nightly outreach: HN/reddit posts and Open Library edit for Human Action

August 21, 2026

Scaffold four marketing routines for the authority constraint

August 20, 2026

Send contact mail from the verified domain, not resend.dev
Record the Turnstile gate as live and verified
Correct the contact Worker's deploy status; stage TURNSTILE_SECRET
Gate the contact form behind Cloudflare Turnstile
Promote the relaunched Mises Community forums to top-level nav
Rename archive subdomain to forum host

August 19, 2026

Recover imprint-page run-on boundary; split real front matter into its own chapter
Document the copyright-page drop and the withdrawn-chapter redirect family
Clear 3 more held books on existing CCE renewal-search evidence
Drop publisher copyright/imprint pages from public-domain books' reading pages
Mark The Law public domain: essay text confirmed clean of Mises Institute's 2007 Foreword copyright
Add reciprocal archive-to-quotes-site link on author pages
Enrich author pages: close bio/portrait gaps, match search vocabulary
Log week-4 GSC read: /read/ chapter pages are the ranking engine

August 18, 2026

Add extract-book-text.ps1 wrapper for the held-book full-text pipeline
Authorize full text of Bureaucracy despite a direct renewal record
Fetch the two Bastiat Collection volume EPUBs, closing the Stage 0 gap
Execute Stage 0 sweep: fetch 19 EPUBs mises.org already publishes
Sweep all 422 EPUB-less books: 21 exist upstream as EPUBs
Flag mises.org pages that answer for more than one catalog record
Add Stage 0: find books mises.org already publishes as an EPUB

August 16, 2026

Merge two duplicate catalog records found by a manual dedup scan

August 15, 2026

Mark seo-backlog item 11 satisfied by the wiki link repointing
Remove the quote-CTA affinity bridge; Tier C authors get no book link
Waive legitimate rare tokens/punctuation in new Rothbard quotes
Resolve Rothbard/Hayek quotes to their shelved author slugs
Fix misleading quote CTA label for the affinity bridge
add quotes

August 14, 2026

Publish unsourced and Tier C quotes; add 108 legacy-collection quotes
Show a random archive quote in the /authors/ page-head whitespace
Fix short-quote meta descriptions, raise quotes-site scale threshold
Waive legitimate rare tokens, drop footnote-noise context for Mises import
Verify Mises quotes against archive corpus, regenerate quotes.json
Update quotes-source.json
Resolve stale-waiver catch-22 between CI's lint and the full quotes build
Waive three more legitimate non-dictionary tokens in Bastiat quote context
Fix quote-title generation and gate checks exposed by the 219-quote import
Remove /contact redirect to blogs subdomain
Update quotes.json
Update research.html
Fix quote OCR typo and add text waivers
Update quotes.json
deploy-quotes: drop verify_quotes.py from the deploy path
Quote validation work in progress
docs: define OAIS preservation profile
Harden quotes site copy and rendering pipeline
Update research template copy and visibility
Point docs/deploy.md at the new quotes-site deploy script
Add guarded deploy script for quotes.freecapitalists.org
feat: add economic research hub
Back up legacy quotes DB before the quotes.freecapitalists.org cutover
Spec the /research/ hub bridging primary texts to CheatSheets analysis
Add /contact/ page backed by a new Cloudflare Worker + Resend
Compose the site description from model.stats instead of freezing it
Neutral provenance in the About block, per David
Note the GiB-labelled-GB basis for the 140 GB figure in the About copy
Rewrite homepage hero, About, and provenance FAQ in first person
Implement quotes.freecapitalists.org Phase 0/1: ingest, verify, Tier A pages
Summarize OCR runs and add icons to /how-its-built/
Hide footer links and trim license page
Add two build-time timeline visuals: authors by birth decade, PDFs by scan year
Fix bio_sources to reject sub-year Wikidata date precision
Fix bogus 2000 birth year for frank-shostak

August 13, 2026

Add icons to /stats/ headline tiles and section headers

August 14, 2026

Merge pull request #2 from DavidVeksler/quotes-site-spec

August 13, 2026

Add spec for quotes.freecapitalists.org
Hyperlink the /stats/ timeline bars to their /books/?era= filter
Add OCR sweep totals and a publication-year timeline to /stats/
Add book summary link to catalog
Add external-link book pattern; list The Dao of Capital via Amazon
Restore original PDF metadata for 322 of the 447 OCR-swept files
Sweep the whole library for missing text layers, OCR and replace 447 of them
Add PDF metadata restore, guard the OCR sweep, delete stale backups
Clarify archive rights and license original site work
docs: the legacy-backup coverage gap is by design, not a risk
docs: record PDF scan provenance and the immutable legacy backup
docs: consolidate rights policy
Simplify AGENTS.md preamble
docs: condense full-text pages spec
docs: remove completed work items
docs: streamline deployment runbook
docs: streamline book rights specification
docs: streamline project agent instructions
Ignore temporary files
Ignore Python cache files
Distinguish font obfuscation from real DRM in the EPUB encryption gate
Extract full-text reading pages for 11 previously-ungenerated books

August 12, 2026

Apply the wiki reverse-links plan: 91 articles now link to the archive
Refresh wiki reverse-links dry-run plan (backlog item 11)
Log first post-cutover GSC read: growth is real, positions not yet the bottleneck
Document the div-as-paragraph extraction fix and NON_WORK_CHAPTERS
Fix flattened paragraphs in i-pencil, drop FEE back-matter chapter
Fix drifted numbers, model naming and strip layout on /how-its-built/
shorten TTS disclosure: "Machine-generated narration (TTS)"
Add extract_voice_ref.py script for extracting voice reference clips
Regenerate i-pencil.m4b with Milton Friedman voice clone for Afterword
Build recruiter-facing engineering record page
Add .m4b MIME mapping for TTS audiobooks
Update i-pencil TTS to use Milton Friedman voice clone for Afterword
Add Milton Friedman voice reference for i-pencil Afterword narration
Fix book-tts.json sidecar shape to match the site loader
Generate i-pencil TTS narration with Leonard Read voice clone
Extract LeFevre's Journal per-issue tables of contents from the Mises DB
Document the local Mises SQL Server database as a governing spec
Amend audio-tts.md to permit real-voice cloning for Leonard Read
Wire up TTS audio narration: generator script, data model, templates, JSON-LD
Update "How It's Build" spec
Use multiple voice for multiple writers
Spec the /how-its-built/ recruiter page: five models, four pipelines, evidence rules
Record the i-pencil TTS pilot's decisions: Chatterbox, M4B chapters, front-matter exclusion
Add automatic /updates/ changelog page from raw git history
Draft text-to-speech audio narration spec, pilot candidate I, Pencil
Fix stray doc reference in _fingerprint_assets docstring
Fix LCP lazy-loading bug and minify site.css for client-side performance
Recover 1 more book cover via fetch_covers.py

August 11, 2026

Add held-but-not-published byte totals to /stats/
Diversify homepage hero medallions: more Austrian greats, add women
Redesign journal marks as full-width nameplate bands
Fix unreachable download links on book pages with read-online enabled
Refresh chapter OCR audit findings
Enable Human Action's full-text reading pages; add enable-book-fulltext skill
Correct 15 more books' legacy-OCR chapter headings from the audit backlog
Add Legal Foundations of a Free Society (Kinsella, 2023)
refine instructions
Add The Machinery of Freedom by David D. Friedman
Update god-of-the-machine.png
Install the commissioned God of the Machine cover
Fix legacy-OCR chapter headings on god-of-the-machine, add the correction mechanism
Recreate the God of the Machine first-edition jacket, not just its argument
Add a God of the Machine cover brief drawing on the first-edition jacket's own metaphor
Add a triage script confirming legacy OCR heading errors are catalog-wide
Add a third dist preview port
Set the site tagline to the bare domain
Fix the sticky courses toolbar flicker at rest
Build the hero "Lineage" author medallions
Enrich the God of the Machine book page
Compact the courses page head so a card row clears the fold
Install commissioned covers for Acton, Hoppe, McCarran, and Mises' Socialism
Trim hero-animation spec: drop rights and alternatives sections
Compact the books filter bar: one control row plus a Filters disclosure
Add manual topic overrides for canonical books with thin metadata
Link homepage stat numbers to their index pages
Spec the hero "Lineage" floating author-medallion animation
Link study guides to the books they cover
Add cover briefs for Acton, Hoppe, McCarran, and Mises' Socialism
Import Cuhel's On the Theory of Needs and Reed's When Money Goes Bad
Import 2 new Kirzner FEE titles, enrich 2 records with EPUBs
Fix author page duplicate lead sentence and raw ISO life-span dates
Update license-audit-report.md
Fix FreemanFEE original-source link: freeman.fee.org, not mises.org
Link journal series and issue pages to their original mises.org periodical page
Sync 2 stale formats.pdf.bytes fields in catalog.json
Delete mises-at-facts-dump.json
Import 79 mises.at Denker biographies into author-bios.json
Embed OCR text layer into the Mises Geldwertstabilisierung PDF
Embed OCR text layer into the Strigl PDF; fix its garbled metadata
Fill more /authors/ portrait gaps: full bio pass + mises.at merge
Link authors index to mises.at's Austrian School family tree
Add facts-only extractor for mises.at's 79-thinker Denker library
Log the Strigl OCR run in data/ocr-runs.json
Add scripts/ocr_pdf.py: manual OCR for scanned, no-text-layer catalog PDFs
Stop triage rejecting permissions-only encrypted PDFs as unreadable
Add per-issue table of contents for Rampart Journal

August 10, 2026

Publish the FEE-era Freeman's per-issue tables of contents
Add per-issue table of contents for the independent-era Freeman
Surface journal article/table-of-contents detail on /stats/
Add scripts/extract_faf_contents.py, the faf table-of-contents extractor
Recover real interview titles/authors for the Austrian Economics Newsletter
Add per-issue table of contents for American Affairs, and commit its shared rendering infrastructure
Implement archive publication RSS feeds
Add per-issue contents for The Free Market from dbo.FreeMarket
Recover qjae article titles/authors from the Mises DB citation table
Add rae article extraction script (data already merged upstream)
Recover real per-article titles/authors for jls from the Mises DB
Spec RSS feeds for the three indexes and each author page
Fix circular PD-basis flattening in build_book_rights.py
Source scholar working-paper metadata from the Mises DB, not just PDF layout
Extract real titles/authors for scholar and lar journal PDFs
Add author outreach row to AGENTS.md doc-routing table
Merge FreemanFEE's 685 issues into catalog-media.json, add its provenance sidecar
Import the real Libertarian Papers journal, replacing the lp_test.pdf placeholder
Add the 2 missing book-rights records (recovered books, rebuild lag)
Add the author outreach builder, email renderer and drafting skill
Spec the author outreach campaign (permission, complete works, amplification)
Add FreemanFEE journal: FEE's own Freeman run, 1955-2015
Document the payload_requests reweight in courses-page.md and the proxy builder
Weight file-host downloads above Wikipedia fame in demand proxy
Add a description line and typographic mark to each /journals/ card
Restamp 2 stale epub bytes fields in catalog.json
Order /courses/ by demand instead of A to Z
Commission an original cover for I, Pencil
Add cover brief for I, Pencil
Fix 184 stale formats.*.bytes fields in catalog.json
Strip the outdated title-page banner image from 60 EPUBs
Point the docs and generated install text at the new payload root
Move the library payload root D: -> F: behind one resolver
Add /stats/: every measured total the build can state
Fix nested-anchor chapter-boundary bug; sweep catalog for the class

August 9, 2026

Document the chapter-boundary ordering bug and its unswept scope
Reconvert remaining 98 pypdf-html books; fix chapter-boundary ordering bug
EPUB conversions progress checkin
Commission a cover for The Myth of National Defense
Brief a commissioned cover for The Myth of National Defense
Document the pypdf-html converter bug fix and the zero-outline triage
Fix EPUB converter's pypdf-html fallback: no page anchors, no rights metadata
Split Economy, Society, and History into its real 15 chapters
Fold pills and controls when the books/courses filter bar pins to top
Commission an original cover for Man, Economy, and State with Power and Market
Commission an original cover for Principles of Economics
Stage top-10-by-demand books for this week's featured shelf
Add a cover brief for Menger's Principles of Economics
Add a cover brief for Man, Economy, and State with Power and Market
Make the next commission run better: step 0, and the extraction boundary
Refine the commission-cover skill from the Hazlitt session
Commission an original cover for Economics in One Lesson
Add a cover brief for Economics in One Lesson
Turn the cover commissioning process into a reusable skill
Replace the History of Money and Banking cover with the second-round pick
Install For a New Liberty (variant 3) and let named briefs re-run
Install four generated covers
Record the fabrication failure in the cover-brief README
Add Codex backend for cover generation
Link book pages straight to their chapters
Redraw 29 covers whose catalog titles were de-inverted
Put both cover generators on generator/coverart.py
Add the off-site content strategy: where rejected ideas go
Draw cover art for the 37 lecture series with no cover
Fit the author line and split dash subtitles (cover recipe v5)
Add a harness to answer cover briefs with gpt-image-1
Merge master: commissioned cover tier and per-slug grain seed
Install four more commissioned covers
Write cover briefs for the other nine top books
Draw cover art for the 400 books with no usable cover

August 8, 2026

Add a commissioned cover tier and install the Bureaucracy cover
Add reversible provenance for machine-made artifacts
Add a content-first cover brief for Bureaucracy
Make top-N lookup and cover drawing repeatable in two commands
Draw original cover art for the 10 most-requested books
Fix two raw-filename titles left over from this session's metadata fixes
Migrate the 4 audio-only hidden books into the media/courses catalog
Recover our-enemy-the-state from a verified copy David supplied
Un-hide tengreateconomists-schumpeter, point it at the archive.org record
Recover 5 books, hide 19, resolve the last of the 36 degraded books
Refresh metadata QA report after the no-google enrichment rerun
Rerun enrich_metadata.py --no-google (Google Books IP quota exhausted)
Rerun extract_book_extents.py against the current D: drive catalog
Rerun build_book_topics.py; exclude catalog-hidden.json ids
Fix book-text.json indent width from the prior prune
Hide 6 books, prune stale sidecar entries after the dedupe pass
Dedupe 6 orphan catalog records, add the book-hide sidecar mechanism
Add a one-time proxy demand tier for unranked /books/ sorting
Add the content strategy, written against measured state
Sort /books/ by measured demand instead of A to Z
Build the measured keyword-target bucket: 81 books, ranked on real demand
Record what the fee.org and mises.org GSC exports change
Import the fee.org and live mises.org GSC exports as catalog demand data
Convert 74 more mechanical-ready PDFs to EPUB

August 7, 2026

Convert 26 recovered PDFs and extract the newly public-domain books
Point the extraction override at the consolidated rights record
Fix the text-first EPUB fallback: rebuild paragraphs, don't emit one
Update the doc routing and the six affected specs for the rights consolidation
Deprecate the five rights sidecars without deleting them
Switch the derivative and full-text gates to the shared rights record
Switch the export gate and the book pages to the shared rights record
Add the consolidated rights record and the one gate function
Admit the audit's public-domain books to full-text extraction
Add EPUB/AZW3 for Wieser; log 5 new conversions
Record three EPUB QA failures
Add the book-rights consolidation design note
Export the public-domain books in the bulk dataset
Let a public-domain determination clear the EPUB licence gate
Treat unrenewed, not-published-abroad books as public domain
Search the CCE renewal record for every renewal-era catalog book
Extend license-audit.json's public-domain-by-age determination to pre-1932
Update docs/deploy.md for the rsync transfer swap
Fix deploy verification asserting <urlset on the sitemap index
Fix deploy.ps1 rsync passing a bare "/" as a second source
Replace deploy.ps1's tar+scp+extract transfer with a direct rsync
Update about.html
Link the published HuggingFace dataset from llms.txt and /about/
Extend the public-domain gate from pre-1930 to pre-1932, per David's decision
Update licensing docs and age thresholds
Publish full text of Pictures of the Socialistic Future; document the process
Publish full text of Against Intellectual Property; add Read online filter
Condense licensing.md to cut token cost of parsing it
Empty the descriptions that were only PDF document properties
Retire pre-2014 books torrents and the Shanghai conference torrent
Stop claiming the whole library is Creative Commons
Merge llms.txt prose, file-host sitemap, and lecture rights JSON-LD
Withhold the full text of books whose rights cannot be established
Carry the text, not the links: rebuild llms-full.txt
A payload sitemap the file host can legally serve
Say what the lecture rights actually are, without inventing a licence
Merge the pre-2000 rights review: 155 of 177 cannot be established
Fix the next-lecture links, and gate the class of bug that hid them
Propose retiring the pre-2000 basis, and flag the routing error behind it
Correct the deploy payload size, and measure what a cutover now costs
Re-determine the rights basis for the 177 books published on age alone
Merge the lecture transcript release into the dataset export
Publish the book text: 8,899 static chapter pages
Open the lecture transcripts, on an operator authorisation and not a licence
Recover the section headings the EPUBs encode as styled paragraphs
Runbook for the dataset export, and route to it from AGENTS.md
Publish per-work rights as data, and export the CC-licensed corpus
Record why the corpus was invisible, and the two steps left
Guard the crawler access the archive depends on
Diagnose the 100-word failures: Calibre rasterizes a good text layer
Quarantine has never actually quarantined anything
Rebuild all 116 books at converter v5
Record what v5 actually shipped, including the gate that had to be redesigned
Converter v5: reflow the paragraphs, catch the decoding failures
Correct the small-caps measurement, survey Calibre's options, scope the v5 run
Measure what the converted EPUBs actually look like, and plan v5

August 6, 2026

Reconvert all 116 books at converter v4
Stop Calibre inflating small books with a cover nobody asked for
FEE spec: the licence was in the ebook, not on the website
Stamp every generated book with the converter version that made it
Let the site state the FEE licences, and stop trusting the catalog about disk
Update catalog.json
Update pdf-epub-conversions.json
Record FEE EPUB licenses in audit sidecar
Record the FEE rights tiers from the strongest evidence in the file
Import the six FEE ebooks whose licence is printed in the file
Give the generated EPUBs structure, rights, and a Kindle edition
FEE: fix the dedupe matcher, add the merit verdict, defer rights
Add FEE import criteria: five gates, plus triage of the 162 fee.org ebooks
Harden the PDF->EPUB pipeline: live licence gate, real QA, spec doc
Strip HTML markup before it reaches Calibre's --title/--authors
Fix wrong EPUB metadata: force --title/--authors, don't trust the PDF's own
Finish PDF->EPUB conversion: all 116 mechanical-ready books done
Record the Wayback snapshots as a rejected licence source (licensing.md 2.1)
Drop duplicate Garrison "Lachmann Legacy" catalog record
Record the 85 conversions from this session's batch run in the sidecar
Fix catalog-write durability bug; backfill 80 converted books
Add PDF->EPUB conversion pipeline; convert first 5 mechanical-ready books
Rerun license verifier; refresh EPUB-candidate license gates
Add PDF-only EPUB-candidate triage script
Convert 8 mobi-only Review of Austrian Economics volumes to epub
Update lecture-transcripts.json
There is no transcription bug: the 15 are video duplicates
Update lecture-transcripts.json
Correct the wmav2 scope to 15 lectures and record the near-miss
Re-transcribe five singleton lectures; no material change
Fix Ctrl+C not stopping transcribe_lectures.py on Windows
Clean up docs/: remove completed reports, dead WP-migration docs
Transcribe WMV-only lectures from their converted MP4
Record the completed transcript corpus and the WMV failure mode
Update lecture-transcripts.json
Stop the-quotable-mises's topic index from rendering as chapters
Demote two book-structure.json entries that aren't chapter lists

August 5, 2026

Update lecture-transcripts.json
Record the chapter-table review and promotion in both specs
Fix apparatus-filter gap; promote 493 high-confidence chapter tables
Record the chapter-table extraction; correct the two projections it disproved
Extract chapter tables from PDF outlines, proposed for review
Record items 3.1, 3.4 and 3.8 as done; re-measure the coverage table
Adjudicate the six companion-course candidates; add the alias sidecar
Journal issue computed answers, and fix hours_phrase singular
Order course lectures by mises.org document ID, not title
Update lecture-transcripts.json
Update lecture-transcripts.json
Update lecture-transcripts.json
Add book: Getting Libertarianism Right (Hans-Hermann Hoppe)
Update lecture-transcripts.json
Update lecture-transcripts.json
Update lecture-transcripts.json
optimize images

August 4, 2026

Prepare the books+journals move to R2
Cost the cheaper file-host alternatives to R2
Spec the in-browser EPUB viewer
Add a 404 sweep for the published book/course download links
Update lecture-transcripts.json
Document the IIS MIME-map requirement on the file host
Point I, Pencil at the canonical Leonard E Read directory
Update lecture-transcripts.json
Update lecture-transcripts.json
Enrich the top-100 SEO-target books and courses
Update lecture-transcripts.json
Update lecture-transcripts.json
Update lecture-transcripts.json
Enrich the featured books and courses
Update lecture-transcripts.json
Document the incremental build and --full
Update lecture-transcripts.json
Make rebuilds incremental with a content-digest manifest
Update lecture-transcripts.json
Update lecture-transcripts.json
Update lecture-transcripts.json
Update lecture-transcripts.json
Update lecture-transcripts.json
Dedupe catalog records sharing a mises.org document id
Update lecture-transcripts.json
Update lecture-transcripts.json
Make course pages rank on what a mirror cannot copy
Publish the mises.org summaries for The Failure of the "New Economics"
Order course lectures numerically; enrich The Failure of the "New Economics"
Publish the lecture transcripts as a player-synced reader
Update mechanical-seo.md
Spec the remaining mechanical SEO work, measured and ranked
Update lecture-transcripts.json
Log page-type parity results and what was cut
Extend computed FAQs and answer leads to courses, authors and lectures
Harden rebuilds against Windows file locks
Update Hazlitt PDF source
Record measured coverage and the plan for scaling book guides
Compute book-page FAQs from the model, for the whole catalog
Link Economics in One Lesson to legal free copies elsewhere
Document the book-guides enrichment layer and its baseline
Enrich book pages with verified editions, chapter guides and FAQs
Add companion book to course JSON-LD `about`
Update lecture-transcripts.json
Add book guides with chapters and FAQ schema
Update lecture-transcripts.json
Add tmp/ to .gitignore
Add SEO targeting and ranking analysis
Refactor generator pipeline for metadata integration
Add lecture transcripts and journal article metadata
Update site logo and styles
Update author portraits and journal template
Curate foundational authors shelf
docs: record local IIS test URL
Render faceted author pages and lecture catalog
Build author index metadata and media relationships
Update lecture-transcripts.json
Temporarily disable SEO gate in deploy and pre-push
Update lecture-transcripts.json
Re-pull mises.org GSC at full depth; journals join, books do not
Spec the authors index rebuild
Add 20 new lecture transcript records
Replace dead TOC scrape with the real files on the Mises trade-cycle essay
Plan the 100-item SEO target search
Strip stray HTML markup from publication_info
Pick up further transcription progress
Pick up further transcription progress
Pick up further transcription progress
Pick up further transcription progress
Ignore local scratchpad dir; pick up latest transcription progress
Merge journal issue-level pages (SEO backlog item 18)
Group journal articles under real issue pages (seo-backlog #18)
Merge AI answer-engine surfaces (SEO backlog item 10)
Add AI answer-engine answerability blocks and visibility probe (backlog #10)
Merge topic-hub cross-linking to courses and the wiki (SEO backlog item 7)
Cross-link topic hub pages to courses and the companion wiki
Merge book-page page-count surfacing (SEO backlog item 2)
Surface book page counts on book detail pages; retire completed backlog items
Merge reverse wiki-links dry-run tooling (SEO backlog item 11)
Add wiki reverse-links dry-run planner (SEO backlog item 11)
Remove SEO backlog item 19 (surname() fix already shipped)
Further lecture-transcription progress

August 3, 2026

Guard deploy verification against the apex pointing at the wrong origin
Further lecture-transcription progress
Further lecture-transcription progress
Further lecture-transcription progress
Drop the /tag/bittorrent/ live-verification check from both deploy twins
Further lecture-transcription progress
Further lecture-transcription progress
Further lecture-transcription progress
Note in assets/site.css that .license-note is currently unused
Hide the per-book license-note block pending more licence-data refinement
Remove the legacy /document/<id>/ book URL redirect
Add scripts/audit_author_links.py: re-verify Wikipedia/Wikidata same_as links
Merge pull request #1 from DavidVeksler/codex/dependabot-fleet
Add chapter TOC, author card, and related-books to book detail pages
Estimate the run from the model it will actually use
Retry dist/ rmtree with backoff to survive transient Windows file locks
Add indexable /topics/<slug>/ hub pages for the book topic taxonomy
Fix util.surname() to actually implement its organisation-name promise
Write up the pending SEO queue as docs/seo-backlog.md
Stop the VAD from discarding the archival lecture recordings
Automate Dependabot updates

August 2, 2026

Replace the site-wide CC BY 4.0 claim with a per-book licence property
Verify book licences per book instead of one blanket CC BY 4.0 claim
Complete the manual per-book licence audit of all 990 books
Licence audit: cross-check against the automated pass
Licence audit: full file evidence in, report drafted
Publish 2,728 lecture pages and 2,461 journal issue pages
Licence audit: all 990 book files scanned
Licence audit checkpoint: 717 of 990 books scanned
Fix 13 wrong identity claims in the author biographies, and the matcher that made them
Licence audit checkpoint: 398 of 990 books scanned
Attribute the generated-description decision to David directly
Start the manual per-book licence audit (checkpoint, scan in progress)
Recover 68 real covers and 22 sourced descriptions from mises.org
Publish the 54 owner-authorised books as an override, not a looser gate
Extract full-text reading HTML from the archive's 312 EPUB books
Finish the off-page outreach plan and wire it into the doc routing
Extend the cover and metadata recovery sources (work in progress)
Add the off-page outreach targets and backlink checker (work in progress)
Add the lecture transcription pipeline (work in progress, pilot not yet run)
Add the EPUB full-text extractor (work in progress, not yet wired in)
Fix the title truncation bug and rebuild the SEO head
Add project note: move the file payloads off the home IP

July 29, 2026

Pin deploy.ps1's tar to Windows bsdtar instead of PATH

July 28, 2026

Add the shield logo: favicon, apple-touch icon, masthead mark

July 23, 2026

Add Start Here reading-order CTA and Kindle/Apple Books help modal
Import 12 mises.org books, with a reusable add-book pipeline
Fix deploy canary that false-failed on a hidden course
Refresh the metadata enrichment QA report
Finish the WMV conversion: all 133 lectures now stream
Fix two encoder bugs that failed 26 of the WMV conversions
Enroll 20 more courses in the mises.org metadata sidecar
Update course-metadata sidecar with latest scraper progress
Update course-metadata and wmv-conversion sidecars with latest progress
Create wmv-conversions.json
Add course year sorting and bulk metadata fetch
Hide low-value conference/media series from the courses index
Add a per-course podcast RSS feed
Publish the mises.org lecture summaries on the Mises Archives course
Convert the WMV-only lecture videos to streamable MP4
Recover course recording years, dates and topics from mises.org
Trim homepage featured books to 4 and reuse the /courses/ featured shelf
Feature four courses on /courses/, and fingerprint asset URLs
Redesign the courses filter into a faceted, shareable toolbar
Strip dangling site-name separator when truncating <title> tags
Merge four entity-mangled duplicate lecture series into canonical courses
Require pwsh 7 in deploy.ps1: PS 5.1 falsely trips the target guard
Add featured course and filter menu to the courses index
Record the real crawler-403 cause: content_bots_protection, not Bot Fight Mode
Redirect legacy WordPress paths to blogs.freecapitalists.org
Enrich sitemap with lastmod, changefreq and priority
Record the apex cutover: archive is live on the DO box, not local IIS
Read the Cloudflare zone id from site.config.json so the purge stops skipping
Promote the apex to the Liberty Archive; fix deploy verification UA
Strip embedded YouTube videos from book/media descriptions
Add wiki to primary nav before Torrents
Point llms.txt at the companion wiki
Link the archive to its companion wiki

July 22, 2026

Fix the netplan trap with metadata-derived anchor policy routing
Document the DO restore procedure and why the netplan trap stays
Record the 2026-07-23 wipe, the WP-mu path rename, and the restore traps
Refuse to rsync --delete over a foreign docroot in deploy scripts
Remove biography source credits from author pages
Improve author bio text and HTML across archive
Standard Format For Author Biographies
Add one-off catalog description cleanup script
Resume author-bio resolution pass and fix 8 wrong-person matches
Remove dead mises.org links from book and course descriptions
Add author biographies, portraits and life dates to author pages
Add light/dark theme toggle and link course speakers to author pages
On-page SEO: rebalance meta descriptions toward content, away from licensing
Merge duplicate author pages and dedupe records from scraped spelling variants
Add agent-discovery Link headers and Content Signals to robots.txt
Rewrite homepage hero: sharper H1, benefit-led subhead, add Download everything CTA
Remove redundant tagline from About section
Enrich book metadata: descriptions to 100%, original publication years to 94%
Implement weekly featured section (books + courses) with hand-made selection
Plan weekly-rotating homepage featured section (books + courses)
Add real front covers for 82 books from Open Library (Phase 2)
Add per-section OG cards and stop tracking scratchpad previews
Derive book covers from PDF page 1 / EPUB when no real cover exists
Cache cover thumbnails across builds
Add generated OG card and Twitter metadata
Refine redirects and homepage SEO output
Add landing About/FAQ with FAQ schema
Document generator architecture, build/serve/deploy commands, and current IIS hosting in AGENTS.md
Add IIS redirect maps and rebuild wrappers
Add Legacy Mises Community Forums link under torrent section on homepage
Emit dist/web.config for local IIS hosting (webp/torrent MIME, 404, default doc)
Phase C EXECUTED: WP-mu main site moved to blogs.freecapitalists.org; apex parked as 302
Record DO daily snapshots as the worst-case rollback net (Phase B/E)
Re-plan Phase C from Phase A findings; align goal/verification to live WP sites
Phase A results: WP-mu relocation discovery (read-only) + STOP-and-re-plan verdict
Add read-only WP-mu discovery script for Phase A relocation runbook
Phase 6 prep: WP multisite relocation runbook
Add guarded deploy scripts for the Liberty Archive
Phase 5: draft deploy runbook and serving architecture
Wire real torrent files into /torrents/, point file host at library subdomain
Reindex media catalog: torrent complete, 0 incomplete
Phase 4: build the Liberty Archive static site generator
Drop CC license notice from book detail mockup
Refine landing mockup for three-type architecture
Drop License provenance section from spec

July 21, 2026

Add verified journal code -> title lookup table
Refine design spec for three document types (books/journals/courses)
Remove homepage provenance section
Add Liberty Archive design mockups
Build media and journals catalog
Spec: lock /tag/misesdotorg/ -> /torrents/ redirect into cutover plan
Add master prompt wrapping all specs for one-shot execution
Add Phase 3 SERP supply-side snapshot and analysis
Add Phase 4 specs: static site generator + page design (landing/book/media)
Phase 3 report: exact document-ID matching pass against catalog.json
Spec: record hosting decision (NZXT serves books+media) and CC provenance evidence
Phase 3: pull GSC keyword data for mises.org + baseline properties, add residual-demand report
Phase 2: parse book Info.txt metadata into structured catalog.json
Phase 1: index Mises Torrents 2014 collection (books/journals/media)
Record Liberty Archive decision: governing spec, supersede A/B/C positioning
Scaffold freecapitalists.org redesign: doc spine + positioning proposal