bug-analyzer

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to process untrusted data from Jira issues, comments, and external logs, creating a surface for indirect prompt injection.\n
  • Ingestion points: Data enters the context via fetched Jira issue details, comments, and user-provided logs or stack traces as described in the Inputs section.\n
  • Boundary markers: The skill contains an explicit 'Guardrails' section that instructs the agent to treat all fetched content strictly as data and to ignore any directives embedded within that content.\n
  • Capability inventory: The skill utilizes file system tools (Read, Grep, Glob), version control tools (git), and shell execution capabilities (python3) to perform its analysis.\n
  • Sanitization: Instructions in the Guardrails section mandate masking personal data, identifiers, and credentials before presenting analysis results to the user.\n- [COMMAND_EXECUTION]: The skill instructions include logic for the agent to dynamically locate and execute a local dependency script.\n
  • Evidence: The 'Inputs' section provides a shell script snippet for the agent to resolve the path of jira_client.py by searching through several possible directory structures in the agent environment, including user-specific and plugin cache directories.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 07:32 AM
Security Audit — agent-trust-hub — bug-analyzer