🔄 Weekly pulse report: 2026-09-25

reports/2026-09-25.md

@@ -0,0 +1,77 @@
+# 📊 GitHub Pulse Scout — areal-project/AReaL
+
+**Window:** 2026-09-18T06:28:55.352Z → 2026-09-25T06:28:55.352Z **Issues fetched:** 6
+
+## Activity
+
+| Metric            | Count |
+| :---------------- | ----: |
+| 🆕 New issues     |     4 |
+| ✅ Closed issues  |     2 |
+| 📂 Currently open |     4 |
+
+## Response efficiency
+
+| Metric                 | Value |
+| :--------------------- | ----: |
+| Avg close time (hours) | 417.0 |
+| Avg close time (days)  | 17.37 |
+| Closed sample size     |     2 |
+| Unique issue authors   |     5 |
+
+## Top labels
+
+| Label                   | Count |
+| :---------------------- | ----: |
+| `stale`                 |     1 |
+| `call-for-contribution` |     1 |
+| `roadmap`               |     1 |
+
+## 🔥 Hottest issues
+
+- [#1302](https://github.com/areal-project/AReaL/issues/1302) — [Roadmap] 2026 Q2
+  Milestones (💬 **11**)
+- [#1612](https://github.com/areal-project/AReaL/issues/1612) — [Feature] Migrate
+  multimodal vLLM rollouts to token-in/token-out generation (💬 **4**)
+- [#1757](https://github.com/areal-project/AReaL/issues/1757) — ViRL39K loader stores
+  mathruler's "None" sentinel as gold for a row with unbalanced braces; a reply with no
+  answer then scores 0.9 on that row (💬 **0**)
+- [#1742](https://github.com/areal-project/AReaL/issues/1742) — examples/tau2:
+  user-simulator validation errors are reported as "AssistantMessage" (hardcoded class
+  name in pinned fork) (💬 **0**)
+- [#1741](https://github.com/areal-project/AReaL/issues/1741) — [BUG] examples/tau2:
+  \_clean_messages(for_user=True) deletes the user simulator's own tool calls and
+  results, breaking the telecom domain (💬 **0**)
+
+## 🤖 AI Summary
+
+# Pulse summary — last 7 days
+
+### Overall
+
+AReaL shows **steady but thin** momentum with light traffic and a small maintainer
+bandwidth gap. Critical correctness bugs in examples and a stale roadmap discussion
+point to drift between core platform and downstream demos.
+
+### Trends & focus
+
+- Activity is modest: **4 new issues**, 2 closed, and **4 open** with 5 unique voices.
+- Response efficiency lags (**~17 days** to close) and recent bug reports (#1741, #1742,
+  #1757) remain silent after opening.
+- Participation is distributed but shallow; roadmap and call-for-contribution labels
+  signal reliance on community help.
+- Focus split between backend deprecation, multimodal rollout design (#1612), and
+  example-layer correctness.
+
+### Recommendations
+
+- Triage open example bugs (#1741, #1742) and patch the pinned tau2-bench fork or loosen
+  the coupling.
+- Clarify ownership and timelines on the roadmap (#1302) to convert interest into
+  commits.
+- Tighten issue triage SLAs and add regression tests for data-loading and
+  message-cleaning edge cases.
+
+<small>Generated on 2026-09-25T06:28:55.355Z by
+[GH Pulse Scout](https://github.com/oss-infra/gh-pulse-scout), a tool that analyzes
+GitHub issue data to provide insights into project health and trends with AI 💖.</small>