The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches content from an external API to define its operating rules, which can be exploited if the source is compromised.
Ingestion points: The agent is instructed to fetch content from https://guidelines.sh/api/guidelines?type=social (SKILL.md).
Boundary markers: The instructions do not include delimiters or warnings to treat the fetched content strictly as data, leaving the agent vulnerable to embedded commands.
Capability inventory: The agent uses the fetched information to generate social media content for various platforms.
Sanitization: No sanitization, filtering, or validation of the fetched content is performed before it is used by the agent.
[EXTERNAL_DOWNLOADS]: The skill retrieves data from a non-standard external domain.
The source https://guidelines.sh/api/guidelines?type=social is a third-party site not identified as a trusted or well-known service provider.

human-social