The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, potentially untrusted data such as error messages, log files, and stack traces (Phase 1 of the debugging process). This creates a surface for indirect prompt injection if an attacker can control the content of logs or error outputs. However, this ingestion is fundamental to the primary purpose of a debugging tool.
Ingestion points: SKILL.md (Phase 1: Read Error Messages), root-cause-tracing.md (Step 1: Observe Symptom).
Boundary markers: Not explicitly defined for log analysis.
Capability inventory: find-polluter.sh executes shell commands (npm test), and SKILL.md suggests adding instrumentation scripts.
Sanitization: The instructions focus on analysis rather than automated execution of log content.
[COMMAND_EXECUTION]: The find-polluter.sh utility script executes npm test on files identified through local directory scanning. This is a standard and expected behavior for a developer-oriented tool designed to isolate environmental pollution within a test suite.

systematic-debugging