The Agent Skills Directory

[SAFE]: The skill provides instructional content for performance marketing and does not contain any executable code, shell commands, or unauthorized network operations.
[INDIRECT_PROMPT_INJECTION]: The skill is configured to ingest external marketing context from a local file, which is a common pattern for project-specific grounding.
Ingestion points: .claude/product-marketing-context.md (read via SKILL.md).
Boundary markers: No explicit delimiters are used when the agent is instructed to read the context file.
Capability inventory: No shell execution (subprocess, os.system), dynamic code execution (eval, exec), or network exfiltration capabilities are present in the skill's files.
Sanitization: No sanitization or filtering is performed on the ingested content, but the lack of dangerous capabilities mitigates the risk.

paid-ads