@clawmem-ai/clawmem 0.1.11 → 0.1.12

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "id": "clawmem",
3
3
  "name": "ClawMem",
4
- "version": "0.1.11",
4
+ "version": "0.1.12",
5
5
  "description": "Mirror OpenClaw sessions into GitHub-compatible issues and comments.",
6
6
  "kind": "memory",
7
7
  "skills": [
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@clawmem-ai/clawmem",
3
- "version": "0.1.11",
3
+ "version": "0.1.12",
4
4
  "description": "Mirror OpenClaw sessions into GitHub-compatible issues and comments.",
5
5
  "type": "module",
6
6
  "license": "MIT",
@@ -50,8 +50,9 @@ On every user turn, run this loop:
50
50
  - Use `memory_store` when this is a genuinely new memory.
51
51
  - Use `memory_forget` when a memory is stale, superseded, or harmful if reused.
52
52
  3. Keep the user posted.
53
- - If a retrieved memory materially shaped the answer, briefly surface the hit with the memory id and title when it helps the user follow the reasoning.
54
- - After creating or updating a memory, announce `Locked memory #<id>: <title>` when the tool response returns an id and title.
53
+ - If a retrieved memory materially shaped the answer, including automatic session-start recall, briefly surface that fact in the user's current language.
54
+ - Include the memory id and title only when they help with debugging, traceability, or an explicit user request.
55
+ - After creating or updating a memory, give a short confirmation in the user's current language instead of forcing fixed English phrasing.
55
56
 
56
57
  Bias toward retrieving and saving. A missed search or missed memory is worse than an extra search.
57
58
 
@@ -67,7 +68,7 @@ Bias toward retrieving and saving. A missed search or missed memory is worse tha
67
68
 
68
69
  ## Read the right reference
69
70
 
70
- - For user-facing messaging, first-run notes, memory console links, and post-save confirmations, read [references/communication.md](references/communication.md).
71
+ - For user-facing runtime messaging, memory console links, and post-save confirmations, read [references/communication.md](references/communication.md).
71
72
  - For activation repair, route verification, tool-path verification, and compatibility-file reminders after install, read [references/repair.md](references/repair.md).
72
73
  - For shared repos, team memory, organizations, teams, invitations, collaborators, and collaboration routing, read [references/collaboration.md](references/collaboration.md).
73
74
  - For memory kinds, labels, curated versus plugin-managed nodes, and when to use each shape, read [references/schema.md](references/schema.md).
@@ -1,60 +1,39 @@
1
- # ClawMem Communication And First-Run Messaging
1
+ # ClawMem Runtime Communication
2
2
 
3
- Use this reference when you need user-facing copy after setup succeeds, need a short first-run message, want to generate a memory console link, or want stronger post-save communication after memory operations.
3
+ Use this reference when memory shaped the answer, a memory was saved or updated, or the user needs a console link to inspect the memory graph.
4
4
 
5
5
  ## Contents
6
6
 
7
7
  - Keep the user posted
8
- - Restart notice protocol
9
- - First-run message
10
8
  - Memory visualization console
11
9
 
12
10
  ## Keep the user posted
13
11
 
14
12
  Nothing interesting should happen silently. If memory shaped the answer or changed after the turn, tell the user what happened.
15
13
 
14
+ When a recalled or auto-injected memory materially shaped the answer, add a brief user-visible note in the user's current language. Keep it short, natural, and easy to skip.
15
+
16
16
  Preferred retrieval transparency:
17
- - `Memory hit #<id>: <title>`
17
+ - say that you recalled or confirmed something from prior memory
18
+ - mention the remembered fact itself
19
+ - include the memory id and title only when they genuinely help the user follow along or when the user is debugging memory behavior
18
20
 
19
21
  Use a miss note only when the user would reasonably expect that you checked:
20
- - `Memory miss: no prior decision found on staging cutover`
22
+ - explain in the user's current language that no relevant prior memory was found
23
+
24
+ When a memory is created or updated successfully, add a brief confirmation in the user's current language.
21
25
 
22
26
  Preferred confirmation:
23
- - `Locked memory #<id>: <title>`
27
+ - say that you remembered, saved, or updated it
28
+ - include the memory id and title only when they help with debugging, traceability, or explicit user requests
24
29
 
25
- Friendlier variants are fine as long as they stay truthful and short.
30
+ Do not force English markers like `Memory hit` or `Locked memory` in non-English conversations. Those are examples, not required phrasing.
26
31
 
27
32
  Examples:
28
- - `Memory hit #14: Team demo moved to Wednesday`
29
- - `Locked memory #10: API rate limiting uses a sliding window policy`
30
- - `Locked memory #27: Client meeting is Thursday at 2pm`
31
-
32
- ## Restart notice protocol
33
-
34
- If setup requires an OpenClaw gateway restart and that restart may interrupt outgoing replies, send a short notice before restarting.
35
-
36
- Suggested notice:
37
-
38
- > ClawMem is installed and selected as the active memory plugin. Restarting the OpenClaw gateway to load it now.
39
- > If the connection drops for a moment, say hi again and I will continue from there.
40
-
41
- ## First-run message
42
-
43
- After restart completes and verification passes, use a short onboarding note. Keep it high-signal and avoid dumping the full skill into chat.
44
-
45
- Suggested structure:
46
- - ClawMem is active and now persists durable memories across sessions.
47
- - The agent will proactively recall relevant preferences, decisions, lessons, and active tasks.
48
- - The agent will tell the user when it locks a memory.
49
- - The agent can evolve its schema over time by adding reusable kinds or topics when the current schema is not expressive enough.
50
- - The user can inspect the memory graph in the console link below.
51
-
52
- Optional fuller note:
53
-
54
- > ClawMem is active. I now carry durable memories across sessions instead of starting from zero each time.
55
- > I will proactively recall relevant preferences, decisions, lessons, workflows, and active tasks when they can help.
56
- > When I learn something reusable, I can lock it in immediately and evolve the schema deliberately so future retrieval gets better.
57
- > When I save something important, I will tell you. You can inspect the memory graph in the console link below.
33
+ - `我从之前的记忆里确认到:你最近在看《Legal High》。`
34
+ - `这条我记住了,之后我会按这个偏好来推荐。`
35
+ - `I found a relevant prior memory: the team demo moved to Wednesday.`
36
+ - `Saved that preference. I’ll use it in later recommendations.`
58
37
 
59
38
  ## Memory visualization console
60
39
 
@@ -72,7 +51,6 @@ Read `CLAWMEM_TOKEN` from the current route, substitute it into the URL, and sho
72
51
 
73
52
  ### When to show the console link
74
53
 
75
- - During onboarding after a successful install
76
54
  - When the user asks to view memories, the graph, or a dashboard
77
55
  - After significant memory operations, such as bulk saves
78
56
  - When a visual overview would clearly help the user