feature-discovery

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill utilizes the GitHub CLI (gh) to search for and view issue details. Interactions with GitHub are considered safe as it is an established, well-known service.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection (Category 8) as it is designed to ingest and analyze untrusted data from external sources.
  • Ingestion points: The skill reads codebase files, local documentation, and GitHub issue comments via gh issue view (Workflow steps 3 and 4 in SKILL.md).
  • Boundary markers: There are no explicit instructions or delimiters provided to the agent to ignore or isolate potential instructions embedded within the codebase or issue comments.
  • Capability inventory: The skill uses rg, git, find, and gh for data retrieval (Search Defaults in SKILL.md). All identified tools and instructions are restricted to read-only operations, preventing an attacker from triggering destructive actions through injected content.
  • Sanitization: No sanitization or validation of the ingested content is performed before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 07:31 PM
Security Audit — agent-trust-hub — feature-discovery