Development Roadmap 2026

[aronmolnar]

Objectives

• 🤖 Improve AI agent
• 🚀 Make it easier to get started with SysReptor
• 🛡️ Take steps to implement Cyber Resilience Act requirements

⚡ Quick wins¹

• 🛡️ Add documentation to handbook about incident reporting requirements of the cyber resilience act
• 🤖 Allow using the AI chat in notes
• 🤖 Add copy button for prompts, responses and returned code blocks (Add copy button for AI prompts and responses #650)
• 🤖 Improvements of the system prompt
  • Review other agents for quick wins
• 🚀 Enable useful non-resource-intensive plugins by default for cloud installation
  • cyberchef,renderfindings,scanimport
• 🚀 Add AI chat to docs to interact with documentation and to ask questions
• 🚀 Add NextCloud form for submitting design components and reports
• 🚀 Create form for requesting academia licenses
• 🚀 Add FAQs/insights/information about SysReptor cloud
• Project retention end by tag (Unintended delays in project retention #651)
• Bulk create findings from templates (Bulk finding import into a new project - feature request #635)

Short list

• 🤖 AI tool call: Ask the user questions (AI tool call: Ask the user questions #652)
• 🤖 AI skills (Feature Request: Support AI skills in the UI #603)
• Audit trail for sensitive actions: logins, enable/disable admin permissions, permission changes, add/remove user from project (Audit Trail for Sensitive Actions #579) (pro only)
  • Might be only readable in Django admin
• 🤖 AI Agent UI: transparently show changes made by agent and add revert button
• 🤖 Allow the AI to search report content (findings, sections, notes)
• Project management kanban board Feature Request: Kanban-style project/report workflow board #613

Long list

• 🤖 Button at input fields to auto-fill a certain field Instruct AI to fill out a certain field #655
• Send e-mails on notifications (Send mail on notification #499)
  • E.g., on project create, comment, user added to project, user creation
• Move or copy a finding to another project/Export finding (Move or copy a finding to another project while including the images #524)
• Allow public sharing reports with commenting permissions for quality assurance (QA) [Feature request] Add license free ReadOnly+Comment Role #448
• 📐 Improve reptor and make stable release

Backlog

• 🤖 Show AI chat history/recent chats
  • get a pretty title for ChatThreads => summarize via LLM in after_agent?
• 🤖 Model selection in the chat
• 🤖 Integrated vectordb, holding SysReptor documentation and the current project
• 🤖 Allow using the chat outside the project, for interacting with multiple projects
  • Statistics, summaries, etc.
  • Could use vectordb for easier data access
• 🤖 Allow using the chat in personal notes
• 🤖 Allow analysing images (multimodal or multiple models)
  • Probably requires additional model, most important OCR
    • e.g., https://github.com/microsoft/markitdown
  • Paste images/files into the chat as context
  • E.g., also scan outputs
• 🤖 Speech to Text for reporting fields and/or AI input
  • Needs multiple models allowed to be configured
• 🤖 Token limits per user
  • Would probably be a Labs-only feature to allow users working on certifications to use the AI agent
  • Complexity might be in counting tokens
  • Maybe via LiteLLM
• 🤖 Allow cloud customers configure LLMs in the web interface
  • The api token should not be displayed in the frontend, if it is set as environment variable (e.g., if set by us in cloud)
• 🚀 Add more design components to https://sysreptor.com/designs
• 🚀 Add more demo reports
  • SOC report
  • Compliance report
  • Vuln scan report
• 🚀 Create explainer videos
• Deduplicate files by file content hash to reduce storage size and improve backup performance
• Incremental backups for stored files and images for faster backups.
• User management
  • Refactor permission system: configure guest permissions per user instead of global settings (tbd)
  • Permission level dropdown: superuser > regular user (has all guest permissions) > restricted user (configurable guest permissions)
• Incremental/partial spellchecking for increased performance and stability
• Allow writing comments in notes
• Filter findings/sections by assigned user (Feature Request: Filter Assigned Sections/Findings by Assigned User #616)
• Add STRIDE classification of vulnerabilities (Add STRIDE classification of vulnerabilities #487)

Footnotes

1. Effort of <1 day ↩