necl-content-poster

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection because it ingests and processes untrusted data from user drafts and local configuration files without explicit sanitization or boundary markers. * Ingestion points: Reads content from user-provided drafts, ideas, or news items as specified in SKILL.md. It also proactively reads a company-context.md file from the project root or .claude/ directory. * Boundary markers: There are no specific delimiters (e.g., XML tags or triple backticks) defined to isolate user-provided content from the skill's system instructions or to warn the model to ignore embedded instructions. * Capability inventory: The skill itself contains only natural language instructions. However, as an AI agent skill, it executes in an environment with access to file system and command-line tools; malicious instructions in the input could attempt to leverage these capabilities. * Sanitization: The instructions do not include steps to sanitize, escape, or validate the input text before the model processes it.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 01:27 PM