Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent AI video-generation guidance skill that discloses its provider API use and scoped local preference memory.
Before installing, be comfortable with sending prompts and any reference media you choose to the listed video providers, and periodically review ~/video-generation/memory.md or history.md if you do not want project preferences or cost notes retained locally.
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
test -n "$FAL_KEY" && echo "Fal configured"
Never ask users to paste secrets into chat.
## Memory Updates
The skill instructs the agent to update persistent memory after meaningful sessions with model preferences, prompt patterns, and cost/latency history, but does not warn users that behavioral preferences and usage patterns will be stored. Persistent profiling can expose sensitive workflow habits over time and creates privacy risk beyond the immediate session.
The setup instructions direct the agent to create a local workspace and copy a memory file into the user's home directory without any explicit user notice or consent. This creates persistent local state and modifies user data, which can violate user expectations and privacy requirements even if the content is not inherently malicious.
No suspicious patterns detected.