dev-loop-performance-profiler
Pass
Audited by Gen Agent Trust Hub on Jul 8, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection (Category 8).
- Ingestion points: The agent is instructed to ingest user-supplied context including performance symptoms, environment descriptions, and traffic scale logs in
SKILL.md. - Boundary markers: Absent; there are no clear delimiters or instructions for the agent to ignore potentially malicious directions embedded in the ingested performance data.
- Capability inventory: The skill provides instructions for executing powerful diagnostic commands (
perf record,node --prof,jmap) and writing files (heap snapshots) to the local system (SKILL.md). - Sanitization: No validation or sanitization routines are implemented for the external symptoms or metrics data.
- [EXTERNAL_DOWNLOADS]: The skill includes instructions to install various developer tools from official repositories.
- Evidence: References to
pip install py-spy(profiling-tools.md),cargo install flamegraph(references/profiling-tools.md), andnpx clinic(SKILL.md) are standard setup steps for performance analysis. - [COMMAND_EXECUTION]: The skill relies on diagnostic system commands for profiling.
- Evidence: Instructions include the use of
perf,node,python -m cProfile, anddotnet-traceto capture execution data across different runtime environments (SKILL.md,references/profiling-tools.md).
Audit Metadata