writing-fragments

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it reads user-controlled file content back into the session to prevent overwriting.
  • Ingestion points: The agent is instructed to read the existing markdown file from disk before each write operation.
  • Boundary markers: There are no explicit markers used to distinguish between the agent's instructions and the content read from the file.
  • Capability inventory: The skill's capabilities are limited to reading and appending text to local files; it does not perform network operations or execute shell commands.
  • Sanitization: No specific validation or sanitization is performed on the content read from the file.
  • [SAFE]: The skill's use of file paths and read/write operations is transparently declared and necessary for its functionality as a writing tool. It does not attempt to access sensitive system directories or exfiltrate data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 12:52 PM
Security Audit — agent-trust-hub — writing-fragments