Memory Pill
PassAudited by ClawScan on May 10, 2026.
Overview
Memory Pill is an instruction-only workspace memory setup that writes persistent OpenClaw files and changes agent workflow, but those behaviors are disclosed and scoped to the OpenClaw workspace.
Use this skill only if you want OpenClaw to maintain persistent workspace memory and behavioral rules. Before activation, consider backing up ~/.openclaw/workspace, ask to review diffs before writes, and avoid placing secrets or sensitive personal data in memory files.
Findings (3)
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.
Future agent behavior may be influenced by whatever is stored in these memory files, including stale, incorrect, or sensitive project/user context.
The skill intentionally stores task and agent state in persistent files that may be reused across sessions.
Main claw decides. Subagents execute. BRAIN.md holds state.
Review persistent memory files periodically, avoid storing secrets or highly sensitive information, and ask the agent to show proposed changes before saving them.
This can change local OpenClaw workspace files, though the visible commands are scoped to the skill’s stated workspace-memory purpose.
The skill instructs the agent to run local filesystem commands and create workspace directories.
mkdir -p ~/.openclaw/workspace/{projects,people,areas,clients,decisions,skills,resources,tasks,archives,memory/{daily,facts}}Ask the agent to show a plan or diff before applying changes, especially if you already have important files under ~/.openclaw/workspace.
After using the skill, the agent may prefer planning, persistent workflows, and subagent delegation rather than completing all work directly in the main session.
The skill deliberately changes how the agent should execute work by encouraging orchestration and subagent delegation.
Orchestration — Don't do real work in main session
Install only if you want this orchestration style, and override the workflow explicitly when you want the main agent to handle a task directly.