Overkill Memory System
SuspiciousAudited by ClawScan on May 10, 2026.
Overview
This is a powerful persistent memory skill, but it includes broad background automation, cloud/shared memory flows, and transcript analysis that should be reviewed before installation.
Review this carefully before installing. It appears purpose-aligned for an advanced memory system, but you should disable cron jobs, cloud backup, shared ChromaDB, platform posting, and transcript analysis until you have verified what data is stored, where it is sent, and how to review or delete it.
Findings (5)
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, stale, or sensitive memories could influence future agent behavior across sessions.
The skill is designed to store and reuse context across sessions, but the provided artifacts do not clearly define review, deletion, retention, or per-task consent controls for what becomes future context.
It provides persistent, contextual memory across agent sessions with automatic importance weighting, emotional tagging, and value-based retention.
Use only if you want persistent agent memory; configure clear memory review, deletion, retention, and opt-in rules before enabling broad automated storage.
Private notes, preferences, project details, or conversation-derived memories may be shared with other agents or synced to a cloud provider.
The skill discloses shared multi-agent memory and daily cloud backup, but the artifacts do not clearly bound what data is shared, who can read it, how consent works, or how sensitive memories are excluded.
Multi-Agent Support: Shared + private ChromaDB areas per agent ... Cloud Backup: Supermemory integration for cross-device backup ... background sync only (daily backup).
Keep cloud backup and shared memory disabled until you have reviewed the exact data paths, access controls, exclusions, and provider settings.
The agent could continue performing memory maintenance or posting-related workflows after the immediate task, potentially affecting public or shared spaces.
The skill advertises scheduled background activity and public/platform-facing actions, but the artifacts do not show clear per-action approval, rollback, or containment boundaries.
Full Automation: Cron jobs for cross-session messages, platform posts, diary entries, and proactive memory maintenance
Do not enable cron jobs, cross-session messaging, or platform posting unless each action is explicitly reviewed and approved by the user.
If configured carelessly, conversation content could be sent to an unintended model CLI or command.
Using CLI-accessible model commands is purpose-aligned for error analysis, but it is a broad local tool surface and may pass transcript-derived prompts to whichever command is configured.
Set ACC_MODELS to use any CLI-accessible model ... Each command is invoked with the prompt appended as the final argument.
Only set ACC_MODELS to trusted local or provider CLIs, and avoid enabling transcript analysis until you understand where prompts are sent.
Users have less registry-level visibility into what needs to run and where the code came from.
The package has runnable Python/shell files and manual setup instructions, but the registry metadata does not declare an install mechanism, required binaries, or a source/homepage for provenance.
Source: unknown; Homepage: none; No install spec — this is an instruction-only skill; Code file presence 31 code file(s)
Inspect the included scripts before running them, verify the publisher/source, and avoid enabling optional install or cron steps until reviewed.