Root cause of the recurring README-sync failure: 05.1 modified README.md in
earlier steps then ran `git pull`, which aborts on a dirty tree whenever a
concurrent bot commit advanced develop first. Rewrote its commit step as a
fetch → reset --hard → regenerate → commit → push retry loop (README is a pure
function of inventory, so a hard reset never loses data and eliminates the
merge-conflict / --ours/--theirs class entirely).
Also:
- 03.1/03.2/03.3: wrap the single-shot pull --rebase && push in a 5x retry loop.
- 07.1 PR Guardian: continue-on-error (advisory; keeps the comment, drops the
blocking red ❌ on AI false positives) + PR-scoped concurrency.
- 07.2 Markdown Linter: PR/branch-scoped concurrency with cancel-in-progress.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fix: the per-page resource search box with maturity tag-pills stopped
appearing after navigation.instant was enabled (DOMContentLoaded fires only
on first SPA load). Reworked v2_filter.js to init via Material's document$
observable with a DOMContentLoaded fallback, guard double-injection, bind the
clickable-tag delegation once, and add a no-results state. Cache-bust ?v=2.9.12.
Change: removed the duplicated Markdown "## Table of Contents" from all 156 V2
content pages and from v2_optimizer.py — it forced hundreds of links of scroll
on large pages. Replaced by Material's native sticky "On this page" TOC
(removed toc.integrate, added toc.follow). tags.md keeps its TOC (it is an index).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
_get_entry_geo() called entry.get("url","") before the url key was
injected into the dict, so TLD inference always received an empty
string and all 304 geo-eligible entries were silently dropped.
Fix: add url= parameter to _get_entry_geo and pass the dict key from
generate_digest. Also renames España→Spain in DIGEST_CATEGORIES,
GEO_CATEGORIES, TLD map, and v2_optimizer geo_cats list.
news_digest.json now includes Americas/Europe/Spain/Asia-Pacific data
(star-based fallback; Gemini will re-rank on next Monday CI run).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Titles containing ' | ' (e.g. "Neo, Now in the Terminal | Pulumi Blog")
broke the Markdown table renderer because the unescaped pipe was parsed
as a column separator.
Root cause: v2_optimizer.py escaped pipes in the 'why' column but not
in the link title — one-line fix adds `.replace("|", r"\|")` to `t`.
Also corrects the 4 already-rendered broken entries in tech-digest.md.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The rebase loop never attempted a push after self-healing; it looped
back to `git pull --rebase` and hit a fresh conflict every iteration
(5x), then the final bare push failed non-fast-forward.
Root cause: in a rebase `--ours` = remote side (not our bot commit),
so the README was discarded and rebuilt each round while the remote
kept advancing.
Fix:
- Pull latest *before* committing (shrinks conflict window)
- Retry loop now tries push first, then merge-pulls on rejection
- Uses `-X ours` merge (correct semantics: keep bot's README)
- Re-runs updater after each merge so metrics stay consistent
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Reconstructed the missing 2.6.0-2.9.1 entries from GitHub Release notes,
closing the gap where CHANGELOG.md had stalled at 2.5.8 while releases
reached 2.9.x. Changelog is now complete and continuous through 2.9.9.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The in-repo CHANGELOG had drifted (last entry 2.5.8) while GitHub Releases
went to 2.9.7. Backfill the 2.9.2-2.9.8 entries I have full detail on and
flag the 2.6.0-2.9.1 gap. 2.9.8 also covers the existing gh_pushed
'N/A'->null data cleanup (PR #385).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Completes the v2.9.4 source fix by cleaning the existing data. last_checked
is already INTEGER (applied by a pipeline run after v2.9.6), so this diff is
limited to the 281 gh_pushed values. Deterministic / CI-reproducible.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The four logo hero cards use a 100px <img>; the Intelligence Digest card
uses an emoji <div class=hero-badge-icon> that was only ~32px tall, so it
rendered shorter. Give .hero-badge-icon a 100px flex-centered box (and
matching 12px bottom margin) so all five cards are the same height.
Cache-bust bumped to ?v=2.9.7.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
last_checked stored epoch seconds as REAL; SQLite's version-specific
REAL->text float printing rewrote ~every row on dump, so a locally
regenerated inventory.sql diverged from CI's. Normalize last_checked to
int on the entry before both the SQL record and the YAML dump, and make
the column INTEGER. All readers do coarse day-level staleness checks, so
sub-second precision is unused.
Verified: load->save idempotent for inventory.sql AND inventory.yaml, and
local (SQLite 3.40.1) output byte-identical to CI-equivalent python:3.11
(SQLite 3.46.1). The deterministic data regenerates on the next pipeline
save; committing only the code avoids conflicting with pipeline data churn.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
