The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the gcx CLI and go tool pprof to perform performance analysis. These are legitimate tools used for their intended purpose within a performance investigation workflow.
[EXTERNAL_DOWNLOADS]: References and instructions point to the official Grafana GitHub repository for the gcx tool and the debug-with-grafana skill. These are trusted vendor resources.
[INDIRECT_PROMPT_INJECTION]: The skill processes external profiling data (such as function names, metrics, and labels) which represents a theoretical ingestion point for untrusted data. However, the instructions focus on technical interpretation of structured tool output (JSON, DOT, pprof), and the capability inventory is limited to standard performance profiling operations.
Ingestion points: Profiling data fetched via gcx profiles query and metrics metadata from gcx profiles labels in SKILL.md.
Boundary markers: None specified for the data ingestion.
Capability inventory: Execution of gcx and go tool pprof commands in SKILL.md and references/advanced-analysis.md.
Sanitization: Data is processed through the gcx and pprof tools which parse the profiling data into structured formats.

performance-rca