The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data sources, specifically mentioned in Phase 1 as 'captured artifacts (HAR file, log dump, core dump)'.
Ingestion points: Data enters the agent's context through the analysis of user-provided or environment-captured logs, HAR files, and trace data (referenced in SKILL.md).
Boundary markers: No specific delimiters or safety instructions are defined to prevent the agent from following instructions embedded within these external artifacts.
Capability inventory: The skill utilizes bash scripts, curl, playwright, and git bisect commands to automate the diagnosis loop.
Sanitization: The instructions do not specify any validation or sanitization procedures for the ingested data before it is used to influence the agent's diagnostic hypotheses or tool usage.
[DYNAMIC_EXECUTION]: The diagnostic process involves the creation and execution of temporary scripts to establish a feedback loop (Phase 1). This includes 'CLI invocation with a fixture input', 'Headless browser scripts (Playwright / Puppeteer)', and 'Throwaway harnesses' created at runtime.
[COMMAND_EXECUTION]: The skill facilitates the execution of various system commands and developer tools, including curl for HTTP requests, git bisect run for version bisection, and custom bash scripts (using the scripts/hitl-loop.template.sh template) to interact with the environment.

diagnose