The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to the way user input is handled.
Ingestion points: User-provided content is interpolated via the $ARGUMENTS variable at the end of SKILL.md.
Boundary markers: Absent; there are no delimiters separating the skill's instructions from the user's input.
Capability inventory: The agent using this skill is directed to use powerful tools such as agent-browser for web automation and pgTAP for database interactions.
Sanitization: Absent; the skill does not contain instructions to escape or validate the input in $ARGUMENTS.
[EXTERNAL_DOWNLOADS]: The skill suggests the installation of several external helper skills from GitHub repositories.
Details: It provides npx skills add commands for repositories belonging to trusted organizations such as Vercel Labs and Anthropic, as well as well-known community developers.
[NO_CODE]: This skill consists entirely of markdown documentation and does not contain any executable scripts or configuration files that run code directly.

coding-workflow