OpenClaw Optimised

PassAudited by ClawScan on May 1, 2026.

Overview

This instruction-only guide is coherent and benign; its only notable effect is helping users create persistent CLAUDE.md rules that can influence future Claude Code sessions.

This skill appears safe to use as a writing guide. Before following its advice, remember that CLAUDE.md and .claude/rules files can persistently influence Claude Code behavior, so review them for accuracy and avoid including secrets or overly broad instructions.

Findings (1)

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.

Low

#ASI06: Memory and Context Poisoning

What this means

Rules placed in these files can affect future Claude Code sessions or shared projects, so inaccurate or overly broad instructions could continue influencing the agent later.

Why it was flagged

The skill explicitly guides users on persistent Claude context files, including user-wide preferences and auto-loaded project memory.

Skill content

`~/.claude/CLAUDE.md` | All sessions (user prefs) ... `.claude/rules/*.md` | Auto-loaded as project memory

Recommendation

Keep CLAUDE.md content short, project-specific, and reviewed; avoid adding secrets or broad instructions that should not persist across sessions.