The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to read external context files, creating a potential surface for indirect prompt injection if those files contain malicious instructions.
Ingestion points: The agent is instructed to read .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md in the SKILL.md file.
Boundary markers: The skill does not provide boundary markers or explicit instructions to treat the ingested file content strictly as data.
Capability inventory: No dangerous tools (such as shell execution or file writing) are directly invoked within the skill's scripts.
Sanitization: The skill lacks sanitization or validation logic for the content of the external context files.
[EXTERNAL_DOWNLOADS]: The skill provides references to several well-known experimentation and statistical tools.
References: Includes links to calculators from Evan Miller, Optimizely, AB Test Guide, and VWO for sample size and duration estimation.

ab-testing