accelerator-application

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements strict operational protocols that prioritize security hygiene, such as redacting PII and treating external data as untrusted.
  • [SAFE]: Tool usage is limited to the defined 'allowed-tools' in the frontmatter, following the principle of least privilege and preventing unauthorized tool invocation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes founder data to personalize applications, representing a potential ingestion surface.
  • Ingestion points: Ingests context from 'docs/departments/Executive/founder-context.md'.
  • Boundary markers: Lacks explicit textual delimiters but mandates a 'Think Before Coding' loop to assess risk before tool execution.
  • Capability inventory: Utilizes 'run_shell_command', 'write_file', and 'replace' tools for document generation and task management.
  • Sanitization: Specifically instructs the agent to treat external data as hostile and redact secrets/PII from context before sharing with subagents.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 07:35 AM
Security Audit — agent-trust-hub — accelerator-application