context-optimizer

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted conversation history, which creates a potential surface for Indirect Prompt Injection (Category 8). Maliciously crafted messages in the dialogue could attempt to influence the agent's summarization or optimization logic.
  • Ingestion points: The conversation data is ingested through the 'conversation' list parameters in the detection and compression functions defined in SKILL.md.
  • Boundary markers: The provided code snippets do not implement boundary delimiters or specific instructions to ignore embedded commands within the processed messages.
  • Capability inventory: The skill logic includes capabilities for reading files from disk (via '_load_from_disk') and invoking LLM summarization tasks.
  • Sanitization: No explicit sanitization or validation of conversation content is present to mitigate the impact of adversarial input.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 01:58 PM