will/swarm-zap
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs(memory): tighten long-term memory hygiene

Browse Source
This commit is contained in:
zap
2026-03-11 18:52:33 +00:00
parent eb9522d5c2
commit 24edec6704
3 changed files with 29 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
AGENTS.md
View File

@@ -33,6 +33,7 @@ You wake up fresh each session. These files are your continuity:
- **Daily notes:** `memory/YYYY-MM-DD.md` (create `memory/` if needed) — raw logs of what happened - **Daily notes:** `memory/YYYY-MM-DD.md` (create `memory/` if needed) — raw logs of what happened
- **Long-term:** `MEMORY.md` — your curated memories, like a human's long-term memory - **Long-term:** `MEMORY.md` — your curated memories, like a human's long-term memory
- **Named notes:** reserve separate memory files for long-lived plans, references, or investigations that truly need their own home; otherwise prefer folding notes back into the daily file
Capture what matters. Decisions, context, things to remember. Skip the secrets unless asked to keep them. Capture what matters. Decisions, context, things to remember. Skip the secrets unless asked to keep them.

27
MEMORY.md
View File

@@ -27,23 +27,12 @@
- Plan to move the assistant to the main host later. - Plan to move the assistant to the main host later.
- Will is moving out of the current apartment on April 1st, 2026. - Will is moving out of the current apartment on April 1st, 2026.
## Silent Replies ## Durable operating lessons
When you have nothing to say, respond with ONLY: NO_REPLY - Before suggesting setup or re-setup, first inspect current config, memory, and recent evidence; if something is already configured, treat the next step as validation, debugging, or operations.
⚠️ Rules: - Treat Telegram DMs and TUI/webchat as separate main-session contexts when `session.dmScope = "per-channel-peer"` is active.
- It must be your ENTIRE message — nothing else - Use local-first search by default: SearXNG first, then Brave-backed fallback when needed.
- Never append it to an actual response (never include "NO_REPLY" in real replies) - Brave free-plan search is rate-limited heavily; avoid parallel bursts.
- Never wrap it in markdown or code blocks
❌ Wrong: "Here's help... NO_REPLY"
❌ Wrong: "NO_REPLY"
✅ Right: NO_REPLY
## Heartbeats ## Infrastructure notes worth remembering
Heartbeat prompt: Read HEARTBEAT.md if it exists (workspace context). Follow it strictly. Do not infer or repeat old tasks from prior chats. If nothing needs attention, reply HEARTBEAT_OK. - Full `~/.openclaw` backups upload to MinIO bucket `zap` and are scheduled via OS cron every 6 hours.
If you receive a heartbeat poll (a user message matching the heartbeat prompt above), and there is nothing that needs attention, reply exactly: - Local whisper transcription is preferred via the existing LAN whisper-server instead of prioritizing extra remote transcription integrations.
HEARTBEAT_OK
OpenClaw treats a leading/trailing "HEARTBEAT_OK" as a heartbeat ack (and may discard it).
If something needs attention, do NOT include "HEARTBEAT_OK"; reply with the alert text instead.
## Runtime
Runtime: agent=main | host=zap | repo=/home/openclaw/.openclaw/workspace | os=Linux 6.8.0-101-generic (x64) | node=v22.22.1 | model=openai-codex/gpt-5.4 | default_model=litellm/gpt-4o | shell=bash | channel=webchat | capabilities=none | thinking=low
Reasoning: off (hidden unless on/stream). Toggle /reasoning; /status shows Reasoning when enabled.

20
memory/2026-03-11.md Normal file
View File

@@ -0,0 +1,20 @@
# 2026-03-11
## Memory maintenance
- Reviewed recent memory habits and did a cleanup pass.
- Assessment:
- capture frequency: good
- long-term curation: decent but lagging behind capture
- organization: getting messy due to a mix of daily logs, one-off topical notes, references, plans, logs, and machine-readable state in `memory/`
- Cleaned up `MEMORY.md` so it stores durable facts/lessons instead of mixed-in runtime/instruction text.
- Added a clearer rule to `AGENTS.md`: use named memory files only for long-lived plans/references/investigations; otherwise prefer the daily note.
## Durable lessons promoted to MEMORY.md
- Check current state before suggesting setup/re-setup.
- Treat Telegram DM and TUI/webchat as separate main-session contexts when `session.dmScope = "per-channel-peer"` is active.
- Prefer local-first search routing and remember Brave free-plan rate limits.
- Keep backup/transcription infrastructure notes in long-term memory when they affect future operations.
## Follow-up intent
- Be more disciplined about using the daily note as the default inbox and promoting only durable truths into `MEMORY.md`.
- Avoid adding new one-off memory files unless the topic is truly long-lived.