The Agent Skills Directory

[SAFE]: The skill defines a high-level methodology for software diagnosis using common development utilities like git, curl, and test runners within the context of a local development environment. These operations are essential for the skill's stated purpose of debugging.
[COMMAND_EXECUTION]: The skill instructions and the helper script 'scripts/hitl-loop.template.sh' facilitate the execution of tests and diagnostic CLI commands. These are standard procedures for a software engineering assistant and do not demonstrate malicious intent or unauthorized access.
[PROMPT_INJECTION]: The skill presents an inherent surface for indirect prompt injection as it is designed to process untrusted diagnostic data such as user-provided bug reports and logs. This is documented for technical awareness but assessed as safe within the context of its primary debugging function.
Ingestion points: Phases 1 and 2 in SKILL.md process user bug descriptions and diagnostic artifacts.
Boundary markers: No explicit delimiters or specific instructions to ignore embedded commands are used when processing external data.
Capability inventory: The skill utilizes subprocess execution for shell scripts, git operations, and file modifications across all diagnostic phases.
Sanitization: No explicit filtering or validation of external diagnostic content is defined before the agent incorporates it into its workflow.

diagnose