The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions in SKILL.md direct the agent to read external context files such as .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md at runtime.
Ingestion points: SKILL.md (instructions to load and prioritize content from project-specific marketing context files).
Boundary markers: Absent; the skill does not explicitly instruct the agent to ignore potentially malicious instructions inside these external files.
Capability inventory: The skill is primarily instructional markdown and does not bundle scripts with dangerous capabilities like network operations, file writing, or system command execution.
Sanitization: Absent; the agent is expected to process the raw content of the marketing context files as authoritative background for the task.
[SAFE]: All external tool references (e.g., Stripe, Rewardful, Dub.co, PartnerStack) are provided as evaluation examples for the user rather than automated integrations. No hardcoded credentials, obfuscation, or malicious commands were detected across the skill files.

suede-referrals