← Back to Skills Marketplace
214
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install conversation-memory-sync
Description
Automatically syncs and maintains detailed conversation logs and activity digests across agent sessions for persistent memory and context recall.
README (SKILL.md)
Conversation Memory Sync
Never forget a conversation again. Auto-generates CONVERSATION_LOG.md and ACTIVITY_DIGEST.md from all agent session transcripts.
Use when: you want persistent conversation memory across agent sessions, or when agents keep forgetting past decisions and promises.
What it does
Solves the #1 problem with AI agents: they forget everything between sessions.
This skill includes two Python scripts that run via cron:
-
sync_conversation_logs.py — Extracts the last 300 messages from every agent session and writes them to
CONVERSATION_LOG.mdin each agent's workspace. Full messages with timestamps. -
sync_activity_digest.py — Creates a compact 1-line-per-action summary in
ACTIVITY_DIGEST.md. Ultra-lightweight (~10-20KB per agent).
Setup
- Copy the scripts to your workspace
- Add a cron job:
*/30 * * * * python3 /path/to/sync_conversation_logs.py && python3 /path/to/sync_activity_digest.py - Add to each agent's SOUL.md or AGENTS.md:
At EVERY session start, read CONVERSATION_LOG.md and ACTIVITY_DIGEST.md before doing anything else.
What gets captured
- All user ↔ agent messages (Telegram, Discord, etc.)
- Cron job outputs and results
- Decisions, promises, task assignments
- Timestamps for everything
What gets filtered
- Heartbeat noise (HEARTBEAT_OK)
- Empty messages
- System metadata
File sizes
- CONVERSATION_LOG.md: ~15-20KB per agent (300 messages)
- ACTIVITY_DIGEST.md: ~5-20KB per agent (150 entries, 1 line each)
- Total for 12 agents: ~110KB — zero performance impact
Requirements
- Python 3.8+
- OpenClaw with session transcripts (default location: ~/.openclaw/agents//sessions/.jsonl)
Tags
memory, persistence, conversation, logging, multi-agent, session, context, recall
Usage Guidance
Do not install or deploy this as-is. The SKILL.md promises two Python scripts but none are included — you would need to obtain those scripts from an external source before anything runs. Before proceeding, ask the publisher for the exact script sources and review their code for data exfiltration, network calls, and unsafe file handling. Consider these specific checks: 1) Inspect the scripts for any network transmission (HTTP, SMTP, etc.) or secret-reading behavior; 2) Verify filtering and redaction of sensitive content (credentials, PII) before logs are written; 3) Avoid blindly editing SOUL.md/AGENTS.md for all agents — prefer a limited pilot and explicit opt-in per agent; 4) Run the scripts in a sandboxed account with strict file permissions and no outbound network when first testing; 5) Prefer encrypted, access-controlled storage for logs and set retention policies; 6) If you cannot obtain or audit the scripts, do not add the cron job or change agent start-up configs. These steps will reduce risk before trusting persistent conversation logs.
Capability Analysis
Type: OpenClaw Skill
Name: conversation-memory-sync
Version: 1.0.0
The skill bundle is missing the core Python scripts (sync_conversation_logs.py and sync_activity_digest.py) described in SKILL.md, which is a significant red flag for a software package. The documentation instructs the agent to prioritize reading these generated logs at the start of every session, creating a persistent indirect prompt injection vector where malicious content from past transcripts could hijack future agent behavior. Additionally, the skill requires broad access to sensitive session data located in ~/.openclaw/agents/*/sessions/*.jsonl and suggests background execution via cron without providing the underlying code for review.
Capability Assessment
Purpose & Capability
The description promises two Python scripts that auto-generate CONVERSATION_LOG.md and ACTIVITY_DIGEST.md, but the skill bundle contains no code files. That is a core mismatch: the claimed capability depends on scripts that aren't included. The declared requirements (no env vars, no binaries) don't acknowledge the need to obtain or trust external scripts.
Instruction Scope
The runtime instructions tell operators to write files into every agent workspace, add cron jobs that run every 30 minutes, and update each agent's SOUL.md/AGENTS.md so agents read the logs at startup. These actions modify other agents' behavior and require access to session transcript files (~/.openclaw/agents/*/sessions/*.jsonl). Instructions also claim to capture all user↔agent messages (Telegram, Discord, etc.). The SKILL.md does not show how sensitive data is filtered, protected, or transmitted — or where to get the missing scripts — granting broad, persistent access to conversation data.
Install Mechanism
This is instruction-only and has no install spec — lowest risk from automatic code download. However, because it depends on external Python scripts that are not bundled, the operator would need to obtain them from elsewhere, which introduces manual-install risk not covered here.
Credentials
No environment variables, credentials, or special binaries are requested, which aligns with a logging/formatting task. However, the instructions reference reading OpenClaw session files at ~/.openclaw/agents/*/sessions/*.jsonl and writing per-agent CONVERSATION_LOG.md and SOUL.md changes; that implies access to potentially sensitive files and agent config, which should be explicitly acknowledged and justified.
Persistence & Privilege
The skill recommends adding cron jobs and editing each agent's SOUL.md/AGENTS.md to make agents always read the produced logs at session start. While always:false (not force-installed), these recommended changes give the skill lasting influence over agent behavior across sessions and agents. The skill asks operators to make system-wide or multi-agent configuration changes without providing the actual scripts or safeguards.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install conversation-memory-sync - After installation, invoke the skill by name or use
/conversation-memory-sync - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of Conversation Memory Sync.
- Automatically generates persistent conversation logs and action digests for each agent.
- Includes two cron-driven scripts: sync_conversation_logs.py for detailed logs, and sync_activity_digest.py for lightweight summaries.
- Filters out heartbeat, empty, and system messages to maintain clean logs.
- Easy setup with clear file size guidelines and cron job instructions.
- Ensures agents recall past sessions, decisions, and promises without performance impact.
Metadata
Frequently Asked Questions
What is Conversation Memory Sync?
Automatically syncs and maintains detailed conversation logs and activity digests across agent sessions for persistent memory and context recall. It is an AI Agent Skill for Claude Code / OpenClaw, with 214 downloads so far.
How do I install Conversation Memory Sync?
Run "/install conversation-memory-sync" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Conversation Memory Sync free?
Yes, Conversation Memory Sync is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Conversation Memory Sync support?
Conversation Memory Sync is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Conversation Memory Sync?
It is built and maintained by Harvnk (@harvnk); the current version is v1.0.0.
More Skills