Three-Dimensional Memory

Manages AI assistant memory by organizing files across three dimensions: timeline logs, full conversation backups, and topic-based project folders for easy r...

This skill is internally coherent and does what it says: it instructs an assistant to persist daily logs, full conversation transcripts, and topic folders to local files. Before installing, consider the privacy risks: full conversation backups can contain secrets, PII, or sensitive business data. Ask how and where the files will be stored (workspace path), whether they will be encrypted at rest, who/what can read them, how long backups are retained, and whether transcripts should be redacted (credentials, API keys, personal data). If you intend to use this in a shared or cloud-mounted workspace, add encryption, access controls, and a retention/rotation policy, or modify the workflow to redact or exclude secrets from backups. The supplied example script writes plain files to the current workspace — review and run it manually first to confirm it behaves as you expect.

Type: OpenClaw Skill Name: three-dimensional-memory Version: 1.0.0 The skill bundle is primarily benign, focusing on AI memory management through file organization. However, the `examples/daily-workflow.sh` script contains a shell injection vulnerability in the line `ls -1 topic-memory/ 2>/dev/null | while read topic; do echo " 📁 ${topic}/"; done`. If a directory within `topic-memory/` were maliciously named (e.g., `$(rm -rf /)`), this could lead to arbitrary command execution when the script is run. While this is an example script and not indicative of malicious intent by the author, it represents a lack of input sanitization, classifying it as a vulnerability and thus 'suspicious' according to the provided guidelines.