NYC

analyze-performance

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • NO_CODE (SAFE): The skill consists entirely of a markdown documentation file (SKILL.md) and does not include any executable scripts, configuration files, or automated installation logic. It functions as a set of instructions for the agent to follow when interacting with the Sentry MCP server.
  • Indirect Prompt Injection (LOW): The skill defines a workflow for processing external data (Sentry performance metrics). While this represents a potential surface for indirect prompt injection if the Sentry data were maliciously crafted, the skill itself possesses no sensitive capabilities that could be exploited.
  • Ingestion points: Sentry transaction data and performance metrics fetched via the Sentry MCP server.
  • Boundary markers: Absent from the prompt instructions.
  • Capability inventory: None; the skill does not define any executable tools or scripts.
  • Sanitization: No data sanitization or validation logic is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:46 PM