2026-04-28

Source

  • Type: local-file
  • Path: /home/topher/.openclaw/workspace-crash-bot/memory/2026-04-28.md
  • Bytes: 1542
  • Updated: 2026-05-03T01:05:57.960Z

Content

# memory/2026-04-28.md
 
## ESP-Claw Full Backup (M5StickS3)
- Device reconnected: 192.168.0.18 (new DHCP lease after power cycle)
- Full backup completed: 70 files to ~/esp-claw-backup/
- memory/ — 7 files including real memory.md content (not boilerplate)
- router_rules/ — 5 rules
- skills/ — 35 files (skills_list.json + capability .md files)
- scripts/ — builtin/ (16), temp/ (4), user/ (3)
- sessions/ — s_telegram.log (14KB), chat_map, skills.json
- static/ — esp-claw.png
- restore.sh helper script present
- **Device has been learning:** memory.md has real entries for -topher (C++/Linux/Lua), comm badge project (GPIO 11/12), m5sticks3_badge_project, topher_profile
 
## ESP-Claw Docs Key Findings
- esp-claw.com/en/flash — web serial console (Web Serial API; Chrome/Windows known issues)
- Reliable serial: `idf.py monitor` or `minicom` at 115200 8N1
- Console commands: `ask hello`, `ask_once`, `cap list`, `auto reload/rules/last`, `skill --activate <name>`
- Tools require `skill --activate <name>` before model can use them
- "working reply but no agent response" → LLM timeout or router rule routing issue
- File API 302 redirect bug: use /files/ direct paths instead of /api/files/read
 
## sdf badblocks Status
- Log file exists at /tmp/badblocks-sdf-20260427-132339.log (188KB)
- Last log update: Apr 27 14:00 UTC — stuck/frozen at 0.23% after 25h elapsed
- Process may have died during gateway reboot
- Needs restart if still not completing
 
## Com Badge
- -topher has started working on Com Badge settings
 

Notes

  • No related pages yet.