The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a variety of local Python scripts to perform SEO tasks, linting, and reporting.
Evidence: Includes execution of scripts/sxo_analyser.py, scripts/dfs_client.py, scripts/seo_lint.py, scripts/google_client.py, scripts/recommend_store.py, scripts/impact_report.py, scripts/watch.py, and scripts/report_publish.py.
[INDIRECT_PROMPT_INJECTION]: The workflow involves ingesting and processing untrusted data from external sources like live Search Engine Results Pages (SERP) and People Also Ask (PAA) clusters.
Ingestion points: External data is brought into the agent context via the dfs_client.py and google_client.py scripts (referencing SKILL.md Steps 2 and 3).
Boundary markers: The skill provides explicit instructions to the agent to "Separate what was measured from what was inferred" and to avoid inventing emotional states or ranking signals, which serves as a conceptual boundary against being misled by external content.
Capability inventory: The agent has the ability to run local scripts, write analysis files, and persist recommendations to a local store.
Sanitization: The instructions do not specify technical sanitization methods, but emphasize strict reliance on "deterministic baseline" results from the analyzer script.
[EXTERNAL_DOWNLOADS]: The skill connects to external services to retrieve SEO and performance data.
Evidence: The workflow utilizes DataForSEO and Google's official APIs for PageSpeed and GA4 data. These are well-known, legitimate service providers used for their intended analytical purposes.

workflow-sxo