The Agent Skills Directory

[PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted external data (stack traces, logs, and thread dumps), which creates a risk of indirect prompt injection attacks. \n
Ingestion points: Content is read from application.log, thread-dump.txt, and user-provided text snippets as described in the Quick Start and Instructions. \n
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat ingested log data as non-executable text. \n
Capability inventory: The skill is granted access to the Bash tool, as well as Read, Grep, and Glob. \n
Sanitization: There is no evidence of sanitization or filtering of the input data before it is processed by the agent. \n- [COMMAND_EXECUTION]: The skill documentation encourages the use of the Bash tool to run diagnostic commands such as mvn dependency:tree, gradle dependencies, and jstack. While these tools are standard for the described debugging tasks, providing this capability alongside the ingestion of untrusted logs increases the potential impact of an injection attack.

java-best-practices-debug-analyzer