The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructs the agent to read external project context files to define product and audience parameters, which creates a surface for indirect prompt injection if those files contain malicious instructions.
Ingestion points: The skill explicitly asks the agent to read .claude/project-context.md and .cursor/project-context.md for product and audience context.
Boundary markers: None. There are no instructions provided to treat the content of these context files as untrusted or to use delimiters to separate them from the skill's core instructions.
Capability inventory: The skill is informational and focuses on content strategy. It does not request tools for network exfiltration, file system modification, or system-level command execution.
Sanitization: None. The skill does not instruct the agent to sanitize or validate the data ingested from the context files before using it to generate marketing plans.

community-forum