fix: reading-tick dedupe + same-host link preference #77
No reviewers
Labels
No labels
Agent
Compat/Breaking
Kind/Bug
Kind/Documentation
Kind/Enhancement
Kind/Feature
Kind/Security
Kind/Testing
Priority
Critical
Priority
High
Priority
Low
Priority
Medium
Reviewed
Confirmed
Reviewed
Duplicate
Reviewed
Invalid
Reviewed
Won't Fix
Status
Abandoned
Status
Blocked
Status
Need More Info
No milestone
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
joshtronic/igor!77
Loading…
Reference in a new issue
No description provided.
Delete branch "fix/reading-dedupe-and-host-preference"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Two fixes to the discretionary-read executor surfaced by a single
journal log line. Igor (a) re-read thatgirljen.com's "fledgling"
post he'd already journaled about, and (b) on a stovak.net pick
followed the "powered by GitHub Pages" footer link instead of an
actual post.
Dedupe: filter_unread used to grep the URL literally against
log.md. That misses everything Claude wrote in prose ("Read
thatgirljen.com -- fledgling") which never captured the URL. New
rule also matches when the candidate URL's domain and slug appear
on the same line in log.md OR any brain/journal entry. Catches
single-word slugs cleanly; multi-word title slugs ("a-confession-
im-an-ai-first-coder-now") still slip through once, then self-heal
because the harness appends the URL to log.md on success.
Link selection: try_source now prefers same-host candidates. On a
personal blog the homepage links to its own posts; off-host hrefs
are nearly always footer cruft. Falls back to off-host only when
there are no same-host candidates, so aggregator-style sources
keep working.
Adds url_host and url_slug helpers shared by both filters.
Co-Authored-By: Claude Opus 4.7 noreply@anthropic.com
Two fixes to the discretionary-read executor surfaced by a single journal log line. Igor (a) re-read thatgirljen.com's "fledgling" post he'd already journaled about, and (b) on a stovak.net pick followed the "powered by GitHub Pages" footer link instead of an actual post. Dedupe: filter_unread used to grep the URL literally against log.md. That misses everything Claude wrote in prose ("Read thatgirljen.com -- fledgling") which never captured the URL. New rule also matches when the candidate URL's domain and slug appear on the same line in log.md OR any brain/journal entry. Catches single-word slugs cleanly; multi-word title slugs ("a-confession- im-an-ai-first-coder-now") still slip through once, then self-heal because the harness appends the URL to log.md on success. Link selection: try_source now prefers same-host candidates. On a personal blog the homepage links to its own posts; off-host hrefs are nearly always footer cruft. Falls back to off-host only when there are no same-host candidates, so aggregator-style sources keep working. Adds url_host and url_slug helpers shared by both filters. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>