The Agent Skills Directory

[Command Execution] (SAFE): The skill executes the taze CLI tool and suggests running package manager commands (npm, pnpm, yarn). These actions are consistent with the skill's stated purpose of updating project dependencies.
[External Downloads] (LOW): The skill prompts the user to install the taze package globally via npm. While taze is a well-known community tool, installing global packages introduces a dependency on external registries and package authors.
[Indirect Prompt Injection] (LOW): The skill parses output from external command execution (taze).
Ingestion points: Taze output (package names and versions) is processed in SKILL.md (Step 3).
Boundary markers: Absent; the instructions rely on the agent's ability to interpret the tabular output format.
Capability inventory: Command execution of taze with write permissions and suggestion of npm install in SKILL.md and run-taze.sh.
Sanitization: Absent; the skill does not explicitly sanitize package names or versions before processing them.

node-deps