Security audit
Security checks for vulnerabilities and agentic risk
This skill is a coherent guide for making narrated multi-scene videos, with disclosed review gates and no evidence of hidden or destructive behavior.
Before installing, review the `npx skills add` commands and prefer a pinned or preapproved skills CLI/source if your environment requires reproducible installs. Expect this skill to call media-generation services, upload user-provided or generated media/audio, and use ffmpeg locally to assemble outputs.
The skill instructs the agent to install and execute tooling via npx skills add ... without pinning the skills package itself to an exact version or integrity-verified artifact. That creates a supply-chain risk: a future or tampered package release could change behavior and run attacker-controlled code at install/runtime in the agent environment.
This line relies on npx skills without fixing the CLI version, so the command may resolve to whatever package version is current when run. In an agent workflow that may execute installation instructions automatically, this exposes the environment to supply-chain compromise or unexpected behavior changes.
Using an unpinned npx skills command means the package providing skills can drift over time or be replaced upstream, allowing arbitrary code execution during install or invocation. Because the skill is operational guidance for an agent, this is more dangerous than ordinary documentation: it may be followed verbatim in a privileged environment.
The command invokes npx skills without constraining the version of the executable package, creating a classic package-resolution and supply-chain risk. If the resolved package is malicious or unexpectedly updated, the agent may execute untrusted code while trying to satisfy the skill prerequisites.
This installation instruction depends on an unpinned npx-resolved CLI, so the actual code executed is not stable or cryptographically bound. In agent ecosystems, such commands can turn documentation into executable supply-chain exposure with arbitrary code execution potential.
The line instructs use of npx skills with no explicit version pin for the package furnishing that command. This makes the skill non-reproducible and vulnerable to malicious upstream changes, typosquatting, or compromised package publication leading to code execution.
Because npx skills is unpinned, the command may fetch and run a different package version over time, including a compromised release. Given this skill orchestrates media-generation tooling and may run in automated agent environments, the supply-chain risk is credible and material.
The 'install the full suite once' instruction still uses npx skills without pinning the executable package version. That exposes users or agents to upstream package compromise and makes the environment dependent on mutable external state at the moment of execution.
The related-skill installation command again relies on npx skills with no pinned version of the CLI. Repeating this pattern increases attack surface because any of these documented commands could be the path by which an agent fetches and executes a compromised package.
This command is vulnerable for the same reason: npx will resolve and run a mutable package version unless explicitly pinned. In a skill file that may be treated as execution guidance by autonomous tooling, that becomes a practical supply-chain execution risk rather than a merely theoretical documentation issue.
The line instructs dynamic execution of an unpinned CLI package through npx, enabling code execution from a mutable upstream source. If an attacker compromises the package or dependency chain, the agent could execute malicious code while following the skill's setup steps.
An unpinned npx skills invocation can pull different code at different times, defeating reproducibility and opening the door to supply-chain compromise. Because this appears in installation instructions inside a skill, the context makes accidental execution by an automation agent more likely.
This final related-skill install command also uses an unpinned npx skills executable, preserving the same upstream package substitution risk. The broader skill context does not contain overt malicious directives, but it does normalize live installation of external tooling, which increases exposure in automated environments.
No suspicious patterns detected.
