Session Memory
v1.3.0Your agent forgets everything after compaction? This fixes it. Built by the AI Advantage community — the world's leading AI learning platform (aiadvantage.ai...
⭐ 0· 80·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Overall the requested actions (reading session logs, converting to Markdown, building a glossary, analyzing cron jobs) match the stated purpose. Minor inconsistencies: SKILL.md says it scans ~/.openclaw/agents/*/sessions/ but scripts.session-to-memory.py uses a hard-coded ~/.openclaw/agents/main/sessions path; build-glossary.py uses a WORKSPACE env var defaulting to ~/.openclaw/workspace, while session-to-memory writes to ~/.openclaw/workspace/memory/sessions — these are plausible but slightly mismatched and could confuse users running across multiple agents.
Instruction Scope
The runtime instructions are precise and the included scripts implement them. The scripts read local session JSONL files and cron job JSON files and write Markdown under memory/. They do not make network calls or exfiltrate data. However, they do aggregate potentially sensitive session contents (user messages, assistant replies, cron payloads) into memory/ and the SKILL.md assumes these files will be vectorized by the platform's memory search — this increases the attack surface for any sensitive content contained in sessions.
Install Mechanism
No install spec; this is instruction-plus-script content only. Nothing is downloaded from external URLs and no binaries are installed.
Credentials
The skill does not request credentials or environment variables. build-glossary.py optionally respects a WORKSPACE env var (reasonable). Be aware the scripts read files in user home paths (~/.openclaw/...) which is proportional to the stated function but may expose any secrets stored in session logs or cron job payloads.
Persistence & Privilege
always:false and no elevated privileges are requested. The scripts write to memory/ under the workspace (their own data), and read session/cron files. This is appropriate for a memory/indexing tool. They do not modify other skills or global agent configs.
Assessment
This skill appears to do what it claims: convert local session logs to Markdown, build a glossary, and suggest memory-aware cron prompts. Before installing/running: 1) Review the exact paths used by the scripts (session-to-memory.py currently targets ~/.openclaw/agents/main/sessions and writes to ~/.openclaw/workspace/memory); adjust if you store sessions elsewhere or have multiple agents. 2) Recognize privacy risk: session logs and cron payloads frequently contain sensitive data (API keys, personal data, passwords). The scripts will consolidate those into memory/*.md and (per SKILL.md) may be vectorized by memory_search — ensure the vector store and memory files are stored with proper access controls or sanitize secrets first. 3) Test on a copy of your data or a limited agent (use --new or --force cautiously) before enabling cron jobs broadly. 4) If you want to limit scope, run the scripts on a subset of session files or edit the scripts to filter/exclude patterns (e.g., API keys) before indexing. 5) If you need higher assurance, ask the author for explicit documentation of which fields are read and how to opt out of indexing particular sessions.Like a lobster shell, security has layers — review code before you run it.
latestvk974t31m1z8ca2jdcj7f5pzp7d845wph
License
MIT-0
Free to use, modify, and redistribute. No attribution required.