Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Agent Self Reflection 1.0.0
v1.0.0Periodic self-reflection on recent sessions. Analyzes what went well, what went wrong, and writes concise, actionable insights to the appropriate workspace f...
⭐ 0· 220·7 current·7 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description describe periodic analysis of recent sessions; the included script and SKILL.md both only enumerate reading recent sessions (tail -50) and writing structured insights to workspace files (AGENTS.md, TOOLS.md, memory/*, skills/*). No unrelated credentials, binaries, or installs are requested.
Instruction Scope
Instructions are focused and explicit about limiting reads (tail -50) and not loading full JSONL files; they direct the agent to append/edit specific workspace files. Note: the skill necessarily reads session transcripts which may contain sensitive user data — this is expected for the stated purpose but is a privacy consideration rather than an incoherence.
Install Mechanism
No install spec; skill is instruction-only with a small helper script. The script uses local CLI (openclaw), tail, python3 and does not download or execute external code from untrusted URLs.
Credentials
The skill declares no environment variables, credentials, or config paths. The runtime instructions reference only the local OpenClaw sessions directory (~/.openclaw/agents/main/sessions) which is appropriate for reflecting on sessions.
Persistence & Privilege
always:false (no forced inclusion). The skill can be invoked autonomously (normal platform default) and is described to run as a cron job, which is consistent with its purpose and not unusually privileged. It does write to workspace files (its intended output) but does not modify other skills' configs.
Assessment
This skill appears coherent with its purpose, but before installing: (1) confirm you trust the openclaw CLI and the environment the agent runs in, since session transcripts may contain sensitive data; (2) decide and document retention/permissions for memory files (memory/about-user.md can capture PII); (3) run the script manually first to review its outputs (dry-run) and ensure it correctly limits reads to tail -50 as promised; (4) if you will schedule it as a cron job, ensure the job runs with an account that has only the minimal file permissions needed (so it can't access unrelated directories).Like a lobster shell, security has layers — review code before you run it.
latestvk971nnhkevz00y8atbnd4fk7q982wvd0
License
MIT-0
Free to use, modify, and redistribute. No attribution required.