skills/erp-mafia/gnubok/create-ticket/Gen Agent Trust Hub

create-ticket

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes Glob and Grep capabilities to scan the local codebase. This is a core part of its functionality to identify relevant files and line numbers for the 'Where' section of the ticket. These operations are performed within the agent's local environment to gather context.
  • [DATA_EXFILTRATION]: Technical details about the codebase, including specific file paths and line numbers, are sent to an external service (Linear) via the mcp__claude_ai_Linear__save_issue tool. This is the intended outcome of the skill and occurs only after the user has reviewed and explicitly approved the ticket preview.
  • [PROMPT_INJECTION]: The skill implements an indirect prompt injection surface as it ingests and processes untrusted data from the local codebase through its scanning step. Maliciously crafted content within the files being scanned could theoretically attempt to influence the ticket's description or the agent's behavior, although the skill requires a final manual approval step which serves as a mitigation.
  • [SAFE]: The skill demonstrates safe practices by requiring explicit user confirmation ('yes' / 'no' / 'edit') before performing the external write operation to Linear. It also uses a predefined set of internal prompt templates rather than allowing arbitrary file path loading from user input.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 10:52 AM