gs-openspec-to-tickets

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the openspec CLI for validation and listing changes. It also uses the gh (GitHub) and jira-cli tools to interact with issue trackers. The skill includes a specific security practice in tracker-github.md by instructing the agent to write ticket bodies to a file outside the repository to prevent the content from being interpreted as shell syntax.
  • [INDIRECT_PROMPT_INJECTION]: The skill reads and processes the content of several markdown files within the repository to draft vertical slices for tickets.
  • Ingestion points: The agent reads proposal.md, design.md, tasks.md, and multiple spec deltas under the specs/ directory for each change.
  • Boundary markers: Absent. The instructions do not specify any delimiters or ignore-behavior for the content ingested from these files.
  • Capability inventory: The agent has access to shell execution via openspec, gh, and jira-cli, and can use Model Context Protocol (MCP) tools for Linear and Jira. It can create issues, link them, and apply labels.
  • Sanitization: While the skill mitigates command injection by using --body-file for GitHub, it does not describe any sanitization or validation of the ingested text to prevent it from influencing the agent's logic (e.g., through prompt injection instructions embedded in the proposal or design documents).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 04:36 PM