my-first-test-01
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill mostly behaves like a self-improvement logger, but it should be reviewed because it can persistently affect future agent behavior and its package/install identity is inconsistent.
Review the package identity before installing, since the registry name does not match the embedded metadata or install instructions. Use this skill only if you want persistent learning files, approve any promotions into agent memory or prompt files, and keep secrets out of logged errors or transcripts.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A user may think they are reviewing or installing one skill while the documentation points them to a different package or repository.
The package identity, embedded metadata, and recommended install target do not line up, which creates a provenance and wrong-package-install risk.
Registry slug `my-first-test-01` v1.0.2; `_meta.json` slug `self-improving-agent` v3.0.13; SKILL.md recommends `clawdhub install self-improving-agent`.
Align the registry metadata, _meta.json, SKILL.md name, and install commands before installing; avoid following install commands that target a different slug or unpinned source.
Mistaken, sensitive, or adversarial content could become persistent guidance that influences future sessions.
The skill stores and promotes learned content into files that can be reused as future prompt/context, but the artifact does not clearly require explicit approval before promotion or define retention/review controls.
important learnings get promoted to project memory ... OpenClaw injects these files into every session ... SOUL.md # Behavioral guidelines, personality, principles
Require explicit user approval before promoting entries to AGENTS.md, SOUL.md, TOOLS.md, MEMORY.md, CLAUDE.md, or Copilot instructions; periodically audit and prune stored learnings.
If enabled, the hook can add self-improvement reminders to future sessions even when the user did not ask about logging in that session.
The optional OpenClaw hook injects a reminder into bootstrap context whenever enabled; this is disclosed and purpose-aligned, but it persistently nudges agent behavior.
event.context.bootstrapFiles.push({ path: 'SELF_IMPROVEMENT_REMINDER.md', content: REMINDER_CONTENT, virtual: true })Enable the hook only if you want recurring reminders, and disable or remove it if the reminders become intrusive.
Session transcripts or learnings could reveal private project details if shared too broadly.
The skill documents cross-session transcript reading and message sending; it includes explicit-use and sanitization guidance, but the capability can expose sensitive context if used carelessly.
sessions_history — Read another session's transcript ... sessions_send — Send a learning to another session ... Use these only in trusted environments and only when the user explicitly wants cross-session sharing.
Use inter-session tools only when explicitly needed, and share short sanitized summaries instead of raw transcripts or command output.