alterlab-cdm-postproduction

Pass

Audited by Gen Agent Trust Hub on Apr 5, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a workflow that is vulnerable to indirect prompt injection (Category 8). It instructs the agent to ingest potentially attacker-controlled data from the local file system and incorporate it into its decision-making process for network and file operations.
  • Ingestion points: The agent is directed to "read existing project files for context — the screenplay, edit notes, director feedback, or previous cut exports" (SKILL.md, Section: Media Management & Project Setup).
  • Boundary markers: The skill does not provide delimiters or instructions to treat the content of these external files strictly as data, which may allow embedded instructions to override the agent's behavior.
  • Capability inventory: The skill possesses the capability to "search the web" (network access) and "write the deliverable as a properly formatted file" (file system write access) based on the gathered context (SKILL.md, Section: Your Workflow).
  • Sanitization: There are no documented steps for validating or sanitizing the content retrieved from project files before it is used to generate search queries or file content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 5, 2026, 11:44 PM
Security Audit — agent-trust-hub — alterlab-cdm-postproduction