Problem: When developers submit SYSLOG via the 🚀 开发者入口, zhuyuan-issue-reply.yml
intercepts the issue and responds with "已记录。霜砚会在下次巡检时处理" instead of letting
syslog-issue-pipeline.yml handle the full auto-processing loop.
Root causes:
1. zhuyuan-issue-reply.yml triggers on ALL new issues with no exclusion for SYSLOG
2. syslog-issue-pipeline.yml only detects via 'syslog' label which may not exist in repo
3. zhuyuan-issue-reply.js has no guard against SYSLOG issues, falls through to generic fallback
Fixes:
- zhuyuan-issue-reply.yml: Exclude issues with syslog/bingshuo-deploy labels, SYSLOG/系统日志
in title, or '### 广播编号' in body (template field marker)
- syslog-issue-pipeline.yml: Detect SYSLOG issues by label OR title OR body pattern
- zhuyuan-issue-reply.js: Add isSyslog guard in handleIssueTrigger() as defense-in-depth
Co-authored-by: qinfendebingshuo <207279273+qinfendebingshuo@users.noreply.github.com>
The script expected field names (team_status, modules, next_step, waiting_for,
last_synced) that didn't match the actual Notion-synced data (team, module,
current, waiting, last_sync). Added a data normalization step after reading
JSON files to bridge the gap. Also handles nested collaborators.json format
and missing tech_routing/server_info/os fields gracefully.
Co-authored-by: qinfendebingshuo <207279273+qinfendebingshuo@users.noreply.github.com>
copilot-swe-agent[bot]
zhizhi200271