The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external websites and marketing research files which are then used to generate campaign reports and automated prompts.
Ingestion points: The references/intake.md file mandates fetching the brand's live site and reading research files like the customer-words bank and competitor reports.
Boundary markers: There are no specific instructions to use delimiters or safety wrappers when processing this external content.
Capability inventory: The skill possesses significant capabilities, including writing to campaign records (gooseworks:campaign_upsert), saving files to the brain (save-to-brain), and creating scheduled automations (gooseworks:automation_upsert).
Sanitization: The instructions do not define sanitization or filtering logic for the ingested external data.
[DYNAMIC_EXECUTION]: The skill dynamically assembles an agent_prompt for scheduled tasks in references/run-memory.md.
Evidence: The skill uses a template to build a complex prompt string that includes workspace paths and relative file references, which is then passed to the gooseworks:automation_upsert tool to define the behavior of future agent runs.

plan-campaign