slides-grab-plan

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted user data (topics, requirements, feedback) and research results, which are interpolated into prompts for a sub-agent. This represents an indirect prompt injection surface.
  • Ingestion points: User inputs and research results are ingested as defined in SKILL.md and references/plan-workflow-reference.md.
  • Boundary markers: The skill uses simple bracketed placeholders (e.g., [user topic]) within task prompts, offering minimal separation between instructions and untrusted content.
  • Capability inventory: The skill can trigger sub-agent tasks via the Task tool and execute specific commands using the slides-grab CLI.
  • Sanitization: No explicit sanitization or strict boundary markers (such as XML tags or explicit 'ignore' instructions) are present for processing external data.
  • [COMMAND_EXECUTION]: The skill uses the slides-grab CLI tool for style management and previewing. This is a legitimate use of the tool within the skill's defined scope and does not involve risky permissions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 12:01 PM
Security Audit — agent-trust-hub — slides-grab-plan