rails-activity-timeline
Warn
Audited by Gen Agent Trust Hub on Mar 24, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends installing the
raix-railsgem (https://github.com/OlympiaAI/raix-rails) to enable AI summarization features. This dependency originates from an external organization that is not on the trusted vendor list. - [PROMPT_INJECTION]: The AI summarization implementation in
references/ai-summaries.mdis susceptible to indirect prompt injection. - Ingestion points: Data from model fields (
event.field_name,event.from_field_value,event.to_field_value) is interpolated directly into the LLM prompt inActivitySummaryGenerator. - Boundary markers: The implementation uses a system prompt to define the output format, which provides a weak boundary that does not prevent the model from following malicious instructions embedded in the input data.
- Capability inventory: The generated summary is stored in the database (
ActivityEvent#details['summary']) and automatically broadcasted to connected users via Turbo Streams, creating a vector for displaying malicious content or performing downstream attacks. - Sanitization: The input data is truncated to manage length, but the implementation lacks specific sanitization or filtering to neutralize prompt injection payloads within the trackable model's fields.
Audit Metadata