The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx to download and execute the agent-browser utility from the NPM registry and fetches branding data from the vendor-controlled domain branding.fly.dev using curl.
[COMMAND_EXECUTION]: The instructions direct the agent to execute shell commands, including npx, curl, and jq, to interact with branding tools and parse configuration files.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from an external vendor website (https://branding.fly.dev/), which represents a potential injection surface for untrusted content.
Ingestion points: Branding data and tool outputs are retrieved from the external domain in SKILL.md.
Boundary markers: The instructions do not define explicit delimiters or instructions for the agent to ignore potentially malicious content embedded in the fetched assets.
Capability inventory: The skill utilizes shell execution capabilities (npx, curl, jq) across its operational steps.
Sanitization: No content filtering or validation of the remote branding data is implemented.

brand