The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes external processes to handle HTML rendering and quality assurance.
scripts/render_report.sh invokes Google Chrome or Chromium in headless mode to capture screenshots of the generated reports.
scripts/delivery_gate.py and scripts/reconcile_content_diff.py use the subprocess module to manage rendering tasks and browser lifecycle during verification steps.
[DYNAMIC_EXECUTION]: The skill utilizes runtime code loading and script injection techniques.
scripts/reconcile_content_diff.py injects a hardcoded JavaScript payload into HTML files before rendering them in a headless browser to extract human-readable text.
scripts/test_reconcile_content_diff.py uses importlib.util to dynamically load the target script for unit testing purposes.
[INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted data, which creates a surface for indirect prompt injection.
Ingestion points: scripts/reconcile_content_diff.py reads HTML files provided via command-line arguments; assets/regen-docs-template.py ingests Markdown files from a user-configured directory.
Boundary markers: assets/regen-docs-template.py uses specific HTML comments ( / ) to delineate content segments.
Capability inventory: The skill can execute shell commands, manage browser processes, and write to the local file system.
Sanitization: Basic sanitization using html.escape is applied to metadata fields in assets/regen-docs-template.py before they are embedded in the final output.

report-with-html