The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing multiple local Python scripts to retrieve and process SEO data, including scripts/project_memory.py, scripts/recommend_store.py, scripts/drift_store.py, scripts/event_log.py, scripts/impact_report.py, scripts/seo_forecast.py, scripts/seo_lint.py, and scripts/project_dashboard.py.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests data from external sources (website audits) that could contain malicious instructions.
Ingestion points: Data is pulled from local stores via scripts/recommend_store.py, scripts/drift_store.py, and scripts/event_log.py. These stores contain data harvested from external web pages and SEO audits.
Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between the SEO data and the system instructions.
Capability inventory: The skill has the capability to execute shell commands and generate command-line instructions for the user to fix SEO issues (scripts/seo_fix.py).
Sanitization: There is no mention of sanitization or validation of the data retrieved from the local stores before it is presented to the agent or the user.
[PERSISTENCE]: The skill provides instructions for the user to establish persistence using the python scripts/watch.py schedule command, which generates configuration lines for system-level automation tools like cron (Linux) or Task Scheduler (Windows).

seo-briefing