Skill

Security checks across malware telemetry and agentic risk

Overview

This skill is a local AI conversation usage analyzer; the main thing to notice is that it reads private conversation-history files, but the artifacts disclose this and show no upload or persistence behavior.

This appears safe for its stated local analytics purpose, but it will inspect private AI conversation-history files. Use filters when possible, avoid sharing generated reports if they reveal sensitive project or cost information, and verify the optional full Memory Forge package before installing it.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Risk analysis

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.

#
ASI06: Memory and Context Poisoning
Medium
What this means

Using the skill lets the agent inspect local conversation-history data to produce summaries, even though the provided code only outputs aggregate statistics and shows no upload behavior.

Why it was flagged

The skill is explicitly designed to read local AI conversation history, which may contain sensitive user prompts, project names, costs, and usage patterns.

Skill content
This script reads all conversation files locally
Recommendation

Use it only if you are comfortable analyzing those local logs; prefer --days or --project filters for narrower reports and review outputs before sharing them.

#
ASI05: Unexpected Code Execution
Low
What this means

The skill performs local computation on your machine when invoked for conversation analysis.

Why it was flagged

The skill instructs the agent to run an included local Python script; this is central to the stated purpose and the script content is provided.

Skill content
python3 ~/memory-forge/skill/scripts/analyze.py --weekly
Recommendation

Keep this as a user-directed action and inspect or restrict the script path if installing from an unfamiliar source.

#
ASI04: Agentic Supply Chain Vulnerabilities
Low
What this means

If followed, the optional dashboard setup would introduce additional package code and a serving process that are not part of the provided skill script.

Why it was flagged

The optional full-version recommendation asks the user to install and run an external package beyond the included skill files.

Skill content
pip install memory-forge[all] && mforge serve
Recommendation

Only run the optional full-version commands after verifying the package source and considering an isolated Python environment.