The Agent Skills Directory

[SAFE]: The skill contains instructional content for software debugging and system analysis with no malicious intent or code identified.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data such as error logs and stack traces. This creates a surface for indirect prompt injection; however, the impact is minimized by the lack of exploitable actions. 1. Ingestion points: SKILL.md (Review error logs, stack traces, and system behavior). 2. Boundary markers: Absent in the instructions. 3. Capability inventory: None (the allowed-tools field is empty, restricting all tool-based interactions). 4. Sanitization: Absent in instructions.
[METADATA_POISONING]: The frontmatter and description fields are consistent with the skill's stated purpose and do not contain deceptive instructions.

debugger