proposal-customize

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill defines a structured process for accepting user-defined overrides for thesis guidelines. It includes specific boundaries to prevent the removal of core rules like citation consistency or source fabrication.
  • [DATA_EXFILTRATION]: No network tools or external data transfer mechanisms are present. The skill strictly operates on local workspace files and does not access sensitive system paths.
  • [EXTERNAL_DOWNLOADS]: There are no patterns involving the download or reference of external packages, scripts, or remote content.
  • [COMMAND_EXECUTION]: The skill instructions focus on text and TOML configuration editing rather than executing shell commands or spawning subprocesses.
  • [NO_CODE]: The skill consists entirely of markdown instructions and reference documentation without any executable scripts or binary files.
  • [PROMPT_INJECTION]: Indirect injection surface analysis: (1) Ingestion points: User-supplied supervisor requirements enter the agent context via SKILL.md. (2) Boundary markers: The agent is instructed to check for conflicts and explain defaults. (3) Capability inventory: The skill is restricted to writing local files. (4) Sanitization: Instructions mandate declining requirements that contradict hard rules.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 08:18 AM
Security Audit — agent-trust-hub — proposal-customize