Workspace Review

**Judgement:** High-quality, low-risk procedural skill for auditing an OpenClaw workspace against canonical conventions. Well-suited as a default maintenance/health-check skill in agents that use the standard `~/.openclaw/workspace` layout. **Strengths & Benefits** - Covers the full lifecycle of workspace health: structure, file purpose boundaries, memory hygiene, vector indexing, git state, rogue files, file size, and skills layout. - Enforces clear single-purpose roles for core files (`AGENTS.md`, `SOUL.md`, `USER.md`, `IDENTITY.md`, `TOOLS.md`, `MEMORY.md`, etc.), reducing configuration and memory drift. - Encourages safe practices: private workspaces, no public Git pushes, secrets kept outside the workspace, and awareness of automatic memory flushing and token compaction. - Defines a consistent review report format, making it easy to log audits and track changes over time. **Key Risks / Limitations** - Assumes the canonical OpenClaw workspace path (`~/.openclaw/workspace`) and a mostly POSIX-like environment (`ls`, `wc`, `git`), which may not fully match customized deployments. - Will flag “issues” in intentionally non-standard or advanced setups (custom file layouts, different naming conventions, non-default memorySearch configs), potentially generating noisy warnings. - Relies on the agent having appropriate permissions to inspect files and run basic shell commands; restricted environments may limit effectiveness. **Recommended Scenarios** - Enable as a core maintenance skill for any agent using the default OpenClaw workspace conventions. - Run on a schedule (e.g., weekly) and after major changes (bootstrap, refactors, new skills added) to catch drift early. - Invoke when users report strange memory behavior, missing context, or confusion about which files are indexed or loaded. - Use during onboarding of new workspaces to bring them into alignment with OpenClaw’s reference conventions before heavy usage.