The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external audit files (e.g., MARKETING-AUDIT.md, SEO-AUDIT.md) as specified in SKILL.md and references/report-workflow.md. While the skill includes safety constraints to prevent the inclusion of private data, it lacks structural protections for the data ingestion process.
Ingestion points: The agent is instructed to discover and read content from multiple source files, including MARKETING-AUDIT.md, LANDING-CRO.md, SEO-AUDIT.md, and others listed in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters (e.g., XML tags or triple quotes) or provide "ignore embedded instructions" warnings when interpreting the content of these external files.
Capability inventory: The skill has the capability to write a compiled report to the file system (MARKETING-REPORT.md).
Sanitization: SKILL.md contains safety constraints that explicitly forbid the inclusion of private context, personal identifiers, credentials, tokens, or API keys in the generated output.

market-report