feat(interests): a second overnight tier for non-lensing papers

.github/workflows/knowledge_board.yml

@@ -18,9 +18,11 @@ on:
       - "bibliography/**"
       - "reading-queue.md"
       - "arxiv-inbox.md"
+      - "arxiv-interests.md"
       - "index.md"
       - "scripts/board.py"
       - "scripts/inbox_actions.py"
+      - "scripts/interests_actions.py"
   schedule:
     - cron: "45 5 * * 1"
   workflow_dispatch:
@@ -69,6 +71,9 @@ jobs:
           ensure queue-intake  "5319e7" "reading queue: file this paper fully into the wiki"
           ensure queue-add     "fbca04" "arXiv inbox: move this paper into the reading queue"
           ensure queue-dismiss "d4c5f9" "arXiv inbox: drop this suggestion"
+          ensure interests-add     "fbca04" "arXiv interests: move this paper into the reading queue"
+          ensure interests-dismiss "d4c5f9" "arXiv interests: drop this suggestion"
+          ensure interests-clear   "b60205" "arXiv interests: clear a whole day's batch"
 
       # The family look (hero, palette, stylesheet) is defined once in the
       # Brain and imported by the renderer — never copied in here.

.github/workflows/queue_actions.yml

@@ -1,13 +1,21 @@
 name: Queue actions
 
-# The dashboard's mechanical per-paper buttons land here. Three of them:
+# The dashboard's mechanical buttons land here — five per-paper, one
+# per-batch. Six of them:
 #
 #   ✅ `queue-read`     marks a reading-queue.md line `DONE <date> — <title>`
 #                       (never deleted — the reading history)
 #   ➕ `queue-add`      moves an arxiv-inbox.md line into its reading-queue
 #                       section, unread
 #   ✖️ `queue-dismiss`  drops an arxiv-inbox.md line — an un-acted suggestion
 #                       is not reading history, and git history holds it
+#   ➕ `interests-add`      the same move out of arxiv-interests.md, routed by
+#                          the line's own `[Topic]` rather than to one section
+#   ✖️ `interests-dismiss`  the same drop, out of arxiv-interests.md
+#   🧹 `interests-clear`    drops a whole day's batch from arxiv-interests.md,
+#                          revealing the next day — the interests list is a
+#                          backlog to walk, not a seven-day timer, so this is
+#                          the button that moves it along
 #
 # Each commits the file it touched, re-renders the dashboard, and closes the
 # issue. 📥 `queue-intake` and 📑 `queue-cite` are NOT handled here — they stay
@@ -35,12 +43,16 @@ concurrency:
 
 jobs:
   queue-action:
-    # One job, three labels. The gate is unchanged: this is a public repo, and
-    # a stranger's issue must never mutate the queue or the inbox.
+    # One job, six labels. The gate is unchanged: this is a public repo, and
+    # a stranger's issue must never mutate the queue, the inbox or the
+    # interests list.
     if: >-
       (contains(github.event.issue.labels.*.name, 'queue-read') ||
        contains(github.event.issue.labels.*.name, 'queue-add') ||
-       contains(github.event.issue.labels.*.name, 'queue-dismiss')) &&
+       contains(github.event.issue.labels.*.name, 'queue-dismiss') ||
+       contains(github.event.issue.labels.*.name, 'interests-add') ||
+       contains(github.event.issue.labels.*.name, 'interests-dismiss') ||
+       contains(github.event.issue.labels.*.name, 'interests-clear')) &&
       contains(fromJSON('["OWNER", "MEMBER", "COLLABORATOR"]'),
                github.event.issue.author_association)
     runs-on: ubuntu-latest
@@ -67,6 +79,18 @@ jobs:
               action=dismiss
               status=$(python scripts/inbox_actions.py dismiss \
                 --body-file /tmp/issue_body.txt) || true ;;
+            *,interests-add,*)
+              action=interests-add
+              status=$(python scripts/interests_actions.py add \
+                --body-file /tmp/issue_body.txt) || true ;;
+            *,interests-dismiss,*)
+              action=interests-dismiss
+              status=$(python scripts/interests_actions.py dismiss \
+                --body-file /tmp/issue_body.txt) || true ;;
+            *,interests-clear,*)
+              action=interests-clear
+              status=$(python scripts/interests_actions.py clear \
+                --body-file /tmp/issue_body.txt) || true ;;
             *)
               action=read
               status=$(python scripts/queue_mark_done.py \
@@ -76,14 +100,15 @@ jobs:
           # normalise so the close step reports rather than mis-closing.
           case "$status" in
             marked|already-done|added|already-there|dismissed|already-gone) ;;
+            already-clear|cleared:*) ;;
             *) status="${status:-error}" ;;
           esac
           echo "action=${action}" >> "$GITHUB_OUTPUT"
           echo "status=${status}" >> "$GITHUB_OUTPUT"
           echo "queue action ${action}: ${status}"
           # Did anything actually change on disk? Cheaper and more honest than
           # inferring it from the status token.
-          if [ -n "$(git status --porcelain -- reading-queue.md arxiv-inbox.md)" ]; then
+          if [ -n "$(git status --porcelain -- reading-queue.md arxiv-inbox.md arxiv-interests.md)" ]; then
             echo "changed=yes" >> "$GITHUB_OUTPUT"
           else
             echo "changed=no" >> "$GITHUB_OUTPUT"
@@ -97,11 +122,14 @@ jobs:
           git config user.name  "github-actions[bot]"
           git config user.email "github-actions[bot]@users.noreply.github.com"
           case "$ACTION" in
-            add)     msg="queue: add inbox paper to the reading queue" ;;
-            dismiss) msg="queue: dismiss inbox paper" ;;
-            *)       msg="queue: mark paper read" ;;
+            add)                msg="queue: add inbox paper to the reading queue" ;;
+            dismiss)            msg="queue: dismiss inbox paper" ;;
+            interests-add)      msg="interests: add paper to the reading queue" ;;
+            interests-dismiss)  msg="interests: dismiss paper" ;;
+            interests-clear)    msg="interests: clear the day's batch" ;;
+            *)                  msg="queue: mark paper read" ;;
           esac
-          git add reading-queue.md arxiv-inbox.md
+          git add reading-queue.md arxiv-inbox.md arxiv-interests.md
           git commit -m "${msg} (#${ISSUE}) [skip ci]"
           for attempt in 1 2 3; do
             if git push; then exit 0; fi
@@ -117,7 +145,14 @@ jobs:
       - name: Close or report on the issue
         env:
           STATUS: ${{ steps.mark.outputs.status }}
+          ACTION: ${{ steps.mark.outputs.action }}
         run: |
+          # Two suggestion tiers share these statuses, so the comment names
+          # the one the human actually tapped.
+          case "$ACTION" in
+            interests-*) tier="arXiv interests list" ;;
+            *)           tier="arXiv inbox" ;;
+          esac
           case "$STATUS" in
             marked)
               gh issue close "$ISSUE" --comment \
@@ -127,19 +162,25 @@ jobs:
                 "Already marked read in reading-queue.md — nothing to change." ;;
             added)
               gh issue close "$ISSUE" --comment \
-                "Added to the reading queue and cleared from the arXiv inbox. Dashboard refresh dispatched." ;;
+                "Added to the reading queue and cleared from the ${tier}. Dashboard refresh dispatched." ;;
             already-there)
               gh issue close "$ISSUE" --comment \
-                "Already in the reading queue — cleared the arXiv inbox line and left the queue alone." ;;
+                "Already in the reading queue — cleared the ${tier} line and left the queue alone." ;;
             dismissed)
               gh issue close "$ISSUE" --comment \
-                "Dropped from the arXiv inbox. Git history still holds the line. Dashboard refresh dispatched." ;;
+                "Dropped from the ${tier}. Git history still holds the line. Dashboard refresh dispatched." ;;
             already-gone)
               gh issue close "$ISSUE" --comment \
-                "Not in the arXiv inbox any more — nothing to change (it may already have lapsed)." ;;
+                "Not in the arXiv inbox or interests list any more — nothing to change (it may already have lapsed or been cleared)." ;;
+            cleared:*)
+              gh issue close "$ISSUE" --comment \
+                "Cleared ${STATUS#cleared:} paper(s) from the arXiv interests list — the next day's batch is now on the dashboard. Git history still holds the batch. Dashboard refresh dispatched." ;;
+            already-clear)
+              gh issue close "$ISSUE" --comment \
+                "That interests batch is already clear — nothing to change." ;;
             no-section)
               gh issue comment "$ISSUE" --body \
-                "Could not process this queue action — the target section does not exist in reading-queue.md. Handle manually; leaving the issue open." ;;
+                "Could not process this queue action — neither the named section nor the \`## Interests\` fallback exists in reading-queue.md. Handle manually; leaving the issue open." ;;
             *)
               gh issue comment "$ISSUE" --body \
                 "Could not process this queue action (status: ${STATUS:-error}) — the paper's line was not found. Handle manually; leaving the issue open." ;;

.github/workflows/queue_filing.yml

@@ -1,7 +1,9 @@
 name: Queue filing
 
 # The dashboard's 📥 `queue-intake` / 📑 `queue-cite` buttons open prefilled
-# issues; this workflow has Claude file the named paper — bibliography entry,
+# issues — from the reading queue or from either suggestion tier
+# (`arxiv-inbox.md`, `arxiv-interests.md`), which the issue's `file:` names;
+# this workflow has Claude file the named paper — bibliography entry,
 # wiki sources section (minimal for cite, full stub for intake, the human's
 # notes folded in), DONE-mark on the queue line — and opens a reviewable PR
 # that closes the issue. Merge stays human.
@@ -97,12 +99,17 @@ jobs:
 
             Read /tmp/issue_body.txt (a dashboard queue request). Extract:
             - `file:` — which file the line lives in: `reading-queue.md`
-              (already queued to read) or `arxiv-inbox.md` (an overnight
-              suggestion, NOT yet in the queue). Absent means reading-queue.md.
+              (already queued to read), or one of the two overnight suggestion
+              tiers, NOT yet in the queue — `arxiv-inbox.md` (strong lensing)
+              and `arxiv-interests.md` (everything else). Absent means
+              reading-queue.md.
             - `section:` — the reading-queue section header,
-            - `line:` — the exact line as written. An arxiv-inbox.md line is
-              prefixed `<YYYY-MM-DD> — `; that date is the inbox timestamp, not
-              part of the title. Strip it before using the title anywhere.
+            - `line:` — the exact line as written. A suggestion-tier line is
+              prefixed `<YYYY-MM-DD> — `, and an arxiv-interests.md one may
+              carry a further `[Topic] ` after that date. Neither is part of
+              the title: strip both before using the title anywhere. (The
+              topic is the same value as `section:`, kept on the line so the
+              file round-trips on its own.)
             - `notes:` — the human's free-text notes; treat as absent if it
               still holds the "(optional — replace this…)" placeholder.
             Read /tmp/issue_label.txt: `queue-intake` (full filing) or
@@ -129,18 +136,18 @@ jobs:
             4. Record it as read, which depends on `file:`:
                - `reading-queue.md` → prefix the exact `line:` inside its
                  `## section` with `DONE <today's date> — ` (never delete it).
-               - `arxiv-inbox.md` → there is no queue line yet. Append
-                 `DONE <today's date> — <title>[ — <ref>]` to the end of the
-                 `## section` in reading-queue.md (the paper is being filed
-                 now, so it enters the queue already read), then delete the
-                 line from arxiv-inbox.md. An inbox line is a suggestion, not
-                 reading history, so deleting it is correct — the DONE line it
-                 becomes is the history.
+               - `arxiv-inbox.md` or `arxiv-interests.md` → there is no queue
+                 line yet. Append `DONE <today's date> — <title>[ — <ref>]` to
+                 the end of the `## section` in reading-queue.md (the paper is
+                 being filed now, so it enters the queue already read), then
+                 delete the line from whichever of those two files `file:`
+                 names. A suggestion line is not reading history, so deleting
+                 it is correct — the DONE line it becomes is the history.
             5. Run `make validate` and `python -m pytest tests/ -q`; fix your
                own filing mistakes until both are green.
 
-            Touch ONLY bibliography/, wiki/, reading-queue.md and
-            arxiv-inbox.md. Do not
+            Touch ONLY bibliography/, wiki/, reading-queue.md,
+            arxiv-inbox.md and arxiv-interests.md. Do not
             commit — the workflow handles git. If the paper cannot be
             identified confidently, write the reason to /tmp/filing_error.txt
             and stop without editing anything.
@@ -164,7 +171,7 @@ jobs:
           git config user.email "github-actions[bot]@users.noreply.github.com"
           branch="queue-filing/issue-${ISSUE}"
           git checkout -b "$branch"
-          git add bibliography/ wiki/ reading-queue.md arxiv-inbox.md
+          git add bibliography/ wiki/ reading-queue.md arxiv-inbox.md arxiv-interests.md
           if git diff --cached --quiet; then
             echo "::error::Claude produced no changes"; exit 1
           fi

README.md

@@ -47,9 +47,12 @@ Self-contained, shared schema:
 
 [`bibliography/`](bibliography/README.md) holds the canonical BibTeX
 metadata every wiki cites against; [`reading-queue.md`](reading-queue.md)
-is what's waiting to be read and filed, and
-[`arxiv-inbox.md`](arxiv-inbox.md) is the overnight suggestions tier in front
-of it. New knowledge updates the metadata
+is what's waiting to be read and filed. Two overnight tiers sit in front of it:
+[`arxiv-inbox.md`](arxiv-inbox.md) (strong lensing, a seven-day timer) and
+[`arxiv-interests.md`](arxiv-interests.md) (everything else — black holes,
+dark matter, galaxy formation, statistics — as one day's ten at a time, a
+backlog you clear a day at a time rather than a timer). New knowledge updates
+the metadata
 and the claim support together, then passes `make validate` (CI-enforced on
 every push).
 

arxiv-interests.md

@@ -0,0 +1,39 @@
+# arXiv interests
+
+The day's ten most relevant arXiv papers that are **not** strong lensing —
+black holes, dark matter, galaxy formation, statistics and inference, and
+whatever else the digest judges worth a look. Strong lensing has its own tier
+(`arxiv-inbox.md`); nothing appears on both.
+
+Format:
+
+- One paper per line: `<YYYY-MM-DD> — [<Topic>] <title>`, optionally followed
+  by ` — <arXiv id or URL>` (the same title/ref convention as
+  `reading-queue.md`, plus the topic).
+- The date is the arXiv **announcement** date the digest ran against, and it is
+  what groups the file into **day batches**.
+- `[<Topic>]` names the `reading-queue.md` section the ➕ button files the paper
+  into — these papers span domains, so unlike the strong-lensing inbox there is
+  no single destination. A topic naming no real section falls back to
+  `## Interests` (`FALLBACK_SECTION` in `scripts/interests_actions.py`).
+- **A backlog, not a timer.** Nothing here lapses. The Dashboard shows the
+  **oldest un-cleared batch only**, and its 🧹 *clear* button drops that whole
+  day and reveals the next — so a fortnight away is a fortnight of batches to
+  cycle through, one tap each, not a fortnight of lost recommendations. This is
+  the one way this list differs from `arxiv-inbox.md`, which lapses after seven
+  days.
+- Each paper carries the same one-tap actions as the inbox: 📄 straight to the
+  PDF, ➕ add to the reading queue, 📥 intake into memory, 📑 make citeable,
+  ✖️ dismiss.
+- A cleared batch is not lost: git history holds it. Same reasoning as the
+  inbox sweep — the never-delete rule in `reading-queue.md` protects *reading
+  history*, and an un-acted suggestion is not history.
+- One `last digest: <YYYY-MM-DD>` line, directly below the `---`, records the
+  last run of the nightly digest — **papers or none** — so an empty list says
+  *which* kind of empty it is: dated today it is a genuinely quiet day, four
+  days stale it means the filing is broken and papers were lost. Replaced,
+  never accumulated.
+- Written by PyAutoMind's `arxiv_interests.yml`. Hand edits are safe but will
+  race a nightly run; prefer the Dashboard buttons.
+
+---

index.md

@@ -37,6 +37,16 @@ the schema in [`wiki/CLAUDE.md`](wiki/CLAUDE.md).
   (`INBOX_WINDOW_DAYS` in `scripts/inbox_actions.py`). Nothing here has been
   chosen yet: a suggestion is not reading history, so a lapsed line is dropped
   rather than DONE-marked, and git history keeps it.
+- [`arxiv-interests.md`](arxiv-interests.md) — the same tier for everything
+  that is **not** strong lensing: black holes, dark matter, galaxy formation,
+  statistics. One line per paper as `<YYYY-MM-DD> — [<Topic>] <title>[ — <ref>]`,
+  where the topic names the reading-queue section ➕ files it into. It differs
+  from the inbox in one way: it is a **day-batched backlog, not a timer**. Each
+  run appends that day's ten, the dashboard shows the **oldest un-cleared batch
+  only**, and one 🧹 *clear* button drops that whole day and reveals the next —
+  so a fortnight away is a fortnight of batches to cycle through, not a
+  fortnight of lost recommendations. Same five per-paper actions as the inbox.
+  Format and transitions live in `scripts/interests_actions.py`.
 - [`reading-queue.md`](reading-queue.md) — the paper queue AND the reading
   history: `## ` section headers group papers by domain, one title per line,
   and a read paper is never deleted — its line gains a `DONE <date> — `

reading-queue.md

@@ -221,6 +221,10 @@ Bayesian Computation in Astronomy: Novel methods for parallel and gradient-free
 Finding reproducible cluster partitions for the k-means algorithm
 Scalable hierarchical BayeSN inference: Investigating dependence of SN Ia host galaxy dust properties on stellar mass and redshift — 2401.08755
 
+## Interests
+NOTE Where an `arxiv-interests.md` paper lands when its `[Topic]` names no
+NOTE section above — the fallback, not a domain. Re-file from here.
+
 ## SETI
 Imagining the Alien: Human Projections and Cognitive Limitations — 2602.07284
 Results of ten years of UCLA SETI searches with the Green Bank Telescope — 2605.05408