The Agent Skills Directory

[SAFE]: The skill utilizes local JavaScript files (analyze-logs.mjs and clear-logs.mjs) to manage and analyze log data stored in a project's .debug directory. The analysis script safely parses log entries as JSON and provides formatted summaries without executing log content or performing network operations.
[PROMPT_INJECTION]: The skill includes an ingestion surface for indirect prompt injection via the $ARGUMENTS placeholder and by reading raw application logs into the agent's context. This is a standard requirement for the skill's debugging functionality.
Ingestion points: SKILL.md ($ARGUMENTS) and .debug/debug.log (via analyze-logs.mjs).
Boundary markers: The analyze-logs.mjs script acts as a parser, which implicitly delimits log fields, but no explicit prompt-level delimiters are used for the user notes.
Capability inventory: The skill can read local files and execute specific maintenance scripts included in the package.
Sanitization: analyze-logs.mjs uses JSON.parse within a try-catch block to handle log entries, effectively filtering out non-JSON content.

debug-reproduced