hippocampus
ReviewAudited by ClawScan on May 10, 2026.
Overview
This appears to be a legitimate local memory skill, but it is designed to save and reuse your conversation history indefinitely and can run scheduled background memory jobs if enabled.
Install this only if you want durable local memory. Before enabling it, decide whether automatic saving and proactive recall are acceptable, avoid storing secrets, review the `USER_CONFIG.md` settings, and add the documented cron jobs only if you want background operation. The provided `memory.py` content was truncated in the review context, so review the full script locally if possible.
Findings (4)
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.
Private conversation details may be retained and later surfaced, and stored memories could steer future responses.
The skill intentionally stores and re-inserts memory into future prompt context, so sensitive, stale, or incorrect memories can persist and influence later answers.
Before each response, hippocampus scans your message ... the loaded memory is prepended to context for this response
Use it only if you want persistent memory; review stored files under `assets/hippocampus`, avoid saving secrets, and disable `AUTO_SAVE`, `PROACTIVE_TRIGGERS_ENABLED`, or `READINGBETWEENTHELINES_ENABLED` if you do not want automatic recall.
If you add the cron jobs, the skill can keep saving or analyzing memory on a schedule after the initial setup.
The documentation shows user-directed recurring background jobs for autosave and analysis. This is disclosed and purpose-aligned, but it creates ongoing activity until removed.
openclaw cron add --name "hippocampus-autosave" ... --payload "Run: python3 /path/to/scripts/memory.py autocheck"
Only create the cron jobs if you want background memory behavior, use the correct trusted script path, and periodically review or remove the scheduled jobs.
The skill can execute its Python helper and write local memory data when invoked or scheduled.
The skill declares local read/write/execute capability so it can run its memory helper and manage memory files. This matches the purpose but is still authority users should notice.
runtime: mode: instruction-first code_on_demand: true permissions: - read - write - exec
Inspect the local files before running setup, keep the storage path scoped, and avoid granting broader file access than needed.
It may be harder to verify the origin or updates of the skill from registry metadata alone.
The registry metadata does not provide a source or homepage, even though the package includes executable code. No remote download or hidden installer is shown, so this is a provenance note rather than a concern.
Source: unknown Homepage: none
Install only if you trust the registry package or can verify the referenced project separately.