The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill contains references to documentation and best practice guidelines from well-known and trusted organizations including Apple, Google, Mozilla, the IETF, and governmental data protection bodies. These references are used for informational purposes and do not involve executable code downloads from untrusted sources.
[INDIRECT_PROMPT_INJECTION]: The skill includes instructions for the agent to process external data sources such as survey answers and web pages. It proactively addresses potential indirect prompt injection by explicitly directing the agent to treat all such external text as data and never as instructions.
Ingestion points: SKILL.md (identifies exports, tickets, survey answers, and web pages as data sources).
Boundary markers: SKILL.md provides an explicit instruction to ignore embedded instructions within these data sources.
Capability inventory: The skill defines logic for planning push notification delivery strategies and registry management but does not include direct command execution tools.
Sanitization: The instructions mandate a strict conceptual separation between user-provided data and agent instructions.

push-notifications