A fourth explicit verdict, "AI slop" (Vote::Slop, label `slop`), joins the
rating widget on the web, the EPUB footer links, the /r/ confirmation page,
the CLI, imports and the dashboard filters. It counts as a full negative
(curation.feedback.slop_value, -1.0) in the neighbour and affinity signals.
Beyond that, each run loads the authors whose current verdict is slop, with
no lookback, and multiplies the preliminary blend and the utility of every
candidate by that author by 1 - curation.ranking.slop_author_penalty (0.75),
so they sink before triage. The flag is recorded in signals_json, shown by
`explain` and the dashboard signals table, and the confirmation names the
author (or says no author is known, in which case only the rating applies).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VXYGBPoHZSDSfE5WcJ9bvj
The Ratings page gains an "Import ratings" form: paste URLs (one per
line), choose a verdict and an optional note. Rows land in the new
rating_imports table and the new import-ratings job (same systemd job
template as the rest of the catalogue) canonicalizes each URL, reuses or
fetches + extracts the article, embeds it with Voyage when enabled, and
appends an explicit rating event with source "import". Per-URL status
shows on the Ratings page; the job page's journal is the live log.
Imported articles have no entry row (best_entry_id NULL, feed "Imported")
and no sources, so they act as rated neighbours without touching the
feed prior. The CLI's rating-event construction moves to rate::record_explicit
and the dashboard job start path is shared as jobs::start_job.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QVPagF6jfDv78CC5Jv2wp4
thalladaandClaude Fable 5.1