Reddi Self Improving Agent
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
Findings (0)
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.
Incorrect, overly broad, sensitive, or adversarial learning entries could influence future agent behavior if promoted without review.
The skill intentionally stores learnings in files that can become persistent context or instructions for future agent sessions.
OpenClaw injects these files into every session: ... AGENTS.md ... SOUL.md ... TOOLS.md ... MEMORY.md ... .learnings/ ... Broadly applicable learning | Promote to CLAUDE.md, AGENTS.md, and/or .github/copilot-instructions.md
Review and redact learning entries before promotion, and require explicit user approval before editing persistent prompt or memory files.
Future sessions may receive self-improvement reminders and may spend time logging learnings even when the user did not ask in that exact session.
When the optional hook is enabled, it persists as a bootstrap-time context reminder in later sessions.
event.context.bootstrapFiles.push({ path: 'SELF_IMPROVEMENT_REMINDER.md', content: REMINDER_CONTENT, virtual: true })Enable the hook only if you want persistent reminders, and disable or remove it if it becomes intrusive.
If used casually, sensitive context from one session could be shared with another session, or background work could be started without clear boundaries.
The integration guide documents cross-session transcript reads, session messaging, and spawning sub-agents.
sessions_history(sessionKey="session-id", limit=50) ... sessions_send(sessionKey="session-id", message="Learning: API requires X-Custom-Header") ... sessions_spawn(task="Research X and report back", label="research")
Use inter-session tools only with user intent, avoid sending sensitive transcript content, and keep spawned sub-agent tasks clearly scoped.
A manual clone could install code that differs from the reviewed registry artifact.
The manual install instructions reference an unpinned external repository; the registry metadata also provides no homepage/source, although no install spec auto-executes this command.
git clone https://github.com/peterskoett/self-improving-agent.git ~/.openclaw/skills/self-improving-agent
Prefer the reviewed registry artifact or pin and inspect a specific commit before enabling scripts or hooks.
Users may grant or expect Docker availability for a skill whose reviewed behavior does not appear to need it.
Docker is declared as required, but the provided hooks and scripts only output reminders or create local markdown scaffolds, so the dependency is not explained by the included code.
requires: bins: - docker ... security_notes: ... docker is used for containerized learning pipelines.
Confirm Docker is actually needed for your workflow; otherwise treat the requirement as unnecessary and avoid running Docker commands from this skill.