triage-issues

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub issue descriptions, stacktraces, and code examples provided by external users. An attacker could embed instructions within a bug report to attempt to influence the agent's summary or research output.
  • Ingestion points: Issue descriptions and technical content analyzed during Stage 1 (Preview) and Stage 2 (Research) in SKILL.md.
  • Boundary markers: The instructions do not specify explicit delimiters (e.g., triple backticks or XML tags) to isolate the untrusted issue content from the agent's internal reasoning instructions.
  • Capability inventory: The skill is restricted to reading from a local checked-out repository using tools like Read, Glob, and Grep. It lacks capabilities for file writing, system modification, or network exfiltration.
  • Sanitization: The skill does not mention specific filtering or validation steps for the input data before processing.
  • [EXTERNAL_DOWNLOADS]: The skill includes references to external documentation located at clickhouse.com. These are official vendor resources and are documented for human reference rather than automated agent retrieval, following safe practices.
  • [COMMAND_EXECUTION]: The skill instructions specify the use of standard search utilities (grep, glob) to locate modules within the repository. These operations are performed on the local filesystem and are scoped to a predefined source map, presenting no privilege escalation risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 08:48 PM
Security Audit — agent-trust-hub — triage-issues