Self Reflection 1.1.1

Continuous self-improvement through structured reflection and memory

This skill's behavior (local reflection, reading/writing a memory file, and being triggered by OpenClaw heartbeats) matches its description and requests no credentials, which is good. However, the package as published is instruction-only and repeatedly references a CLI binary (bin/self-reflection) that is not included. Before installing or enabling automated heartbeats you should: - Verify the implementation: ask the publisher for the CLI source or inspect the GitHub repo referenced in the README (https://github.com/hopyky/self-reflection). Review the actual script under bin/self-reflection before running it. - Confirm provenance: the registry ownerId in the package metadata does not match the _meta.json ownerId; confirm you are fetching the repository you expect and from the intended author. - If you will run the git clone/install steps, review the cloned files and the self-reflection script for behaviors like network calls, exec of arbitrary commands, or unexpected file writes before running or symlinking the binary. - If you want to allow autonomous heartbeat invocation, only enable it after verifying the binary's code. If you cannot verify the implementation, treat this package as documentation only and do not add the heartbeat entry or run the described binary. Providing the actual CLI script (or a proper install spec) would raise confidence to "high"; without that, the missing implementation is the primary coherence problem.

Type: OpenClaw Skill Name: self-reflection-1-1-1 Version: 1.0.0 The OpenClaw AgentSkills skill bundle 'self-reflection' is classified as benign. All instructions in SKILL.md and README.md guide the AI agent to execute local commands (`self-reflection check`, `log`, `read`, `stats`, `reset`) that interact with designated local files (`~/workspace/memory/self-review.md`, `~/.openclaw/self-review-state.json`). There is no evidence of prompt injection attempting to subvert the agent's purpose, exfiltrate data, execute arbitrary remote code, establish persistence, or perform any other malicious actions. The required binaries (`jq`, `date`) are standard utilities, and the installation process involves a standard `git clone` and symlinking.