Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/SanMuzZzZz/LuaN1aoAgent/llms.txt

Use this file to discover all available pages before exploring further.

This changelog follows the Keep a Changelog format. Each version links to its corresponding commit on GitHub.

​
v0.6.0 — 2026-03-09

​
Fixed

  • core: Fixed a crash caused by the rich library’s special character parser encountering sequences like [/]. (bf51e6f)

​
v0.5.0 — 2026-03-08

​
Added

  • core: Tool timeout configuration, Shared Bulletin Board for real-time cross-subtask finding exchange, and active hypothesis persistence across context compression cycles. (42975d5)
  • rag: Chunk overlap support to improve retrieval recall. (6d38528)

​
Improved

  • arch / prompts / docs: Comprehensive overhaul of the overall architecture, prompt templates, and README documentation. (c2bfe38)

​
v0.4.0 — 2026-03-07

​
Added

  • prompts: Updated prompt content. (6e69331)
  • prompts / core: CTF-specific optimizations and focus areas; updated node type descriptions and execution principles. (cf9a4a9)
  • core: Subtask ID support; improved causal graph command handling and node update timestamp logic. (402c87f)
  • web: Updated audit status display logic; optimized status color handling with support for additional status values. (362a7b9)
  • core / web: New task state management system; additional audit status values; configuration options; frontend display and backend logic refinements. (7a741a1)

​
v0.3.0 — 2026-03-06

​
Fixed

  • agent: Improved process termination logic; added task state management; updated frontend task status display. (7843018)

​
v0.2.0 — 2026-02-28

​
Added

  • prompts / core: Executor and Planner prompt templates with detailed schema definitions and penetration testing workflow principles. (fd682fb)

​
Fixed

​
Docs

​
v0.1.0 — 2026-02-26

​
Added

  • core / tools: Enhanced tool chain, resource governance, and cognitive reasoning framework. (3a590bc)

​
Older versions

​
Fixed

  • rag: Fixed relative import path resolution failure. (9b14ad2)
  • tools: Fixed LLMClient import compatibility issue (two separate commits). (6a1d068, cd4ac83)

​
Added

  • web: Left and right sidebar collapse functionality. (8328ee6)

​
Fixed

  • web: XSS vulnerability — LLM-returned content containing raw HTML tags was being rendered unescaped. Fixed via PR #1. (a1cf830)

​
Fixed

  • app: Root node and execution chain node collapse support. (7ee7cee)

​
Docs

  • readme: Updated installation steps; removed stale links. (b9cf3dc)

​
Added

  • server: Operation logging. (a365535)
  • agent: Session state updates and error handling. (3c90d22)
  • model: Custom sort field for sessions. (b280775)

​
Build

  • Added project dependency declarations. (f5de209)

​
Added

  • web: Task management and rendering improvements; task rename and manual sort order support. (8959f8e)

​
Docs

​
Added

  • graph: Placeholder node logic and phase-transition handling. (f7a66fe)
  • ui: Replaced confirmation dialogs with a unified modal component. (30385ca)
  • database: Atomic graph data writes. (ed01082)

​
Style

  • web: Homepage title update. (fe45fa5)

​
Added

  • ops: Drag-to-sort task list. (5706730)
  • graph / ui: User-controlled node expand/collapse; double-click to toggle subtask collapse state. (8993ec6, e399c9b)
  • graph: Transition animation duration scaled to node count for better rendering performance. (cc7b51b)
  • graph: Race condition detection via render task ID tracking. (8032e69)
  • agent: Success node marking; key success steps extracted from subtasks; Reflector output schema update. (99d7fe6)
  • agent: Success path highlighting from Planner-designated nodes; frontend log output improvements. (fb82011)
  • agent: Session record creation in the database to support frontend task refresh. (43c4e57)
  • agent: Successful subtask node marking and success path highlight logic. (2623a5c)
  • ui: New task creation dialog. (7721349)

​
Refactored

  • core: Subtask execution chain persistence mechanism. (3a65744)

​
Added

  • graph: Evidence strength classification and update logic; failure mode rendering with competing hypothesis resolution; Reflector evidence strength judgment principles. (4b8a11c)
  • executor: Failure mode rendering to improve prompt context. (22d113e)
  • reflector: Reject node logic for causal graph management. (2fb8485)
Initial project scaffolding: core graph engine, Web UI, and Agent execution framework.

​
Added

  • core: MCP server management; task completion graph state updates. (ee0cd7e)
  • planner: Human-in-the-Loop review for branch replanning; dependency correction. (ffae7ae)
  • core: Node completion timestamps and highlight logic. (f5933ba)
  • graph: Global task completion state and success path highlighting. (5269517)
  • ui: Task phase banner display. (f021008)
  • exec: Execution path highlighting and UI internationalization. (b39909c)
  • graph: Node sizing by type; current execution path highlight styling. (b265c58)
  • graph: Current execution path highlight; node and edge style improvements. (5140dbd)
  • graph: Node type handling improvements and detail panel display. (a0e161f)
  • executor: Task state management and frontend node rendering. (1fd5cc7)
  • agent: Operation ID passing via Web UI for unified task identification. (6712f49)
  • intervention: Enhanced HITL request handling with database integration. (32b511f)
  • web: Web service host address configurable via environment variable. (9e8efb2)

​
Refactored

  • web: Task deletion and task status display improvements. (0d6179e)

​
Fixed

  • agent: Async call and event handling logic. (ad659ec)
  • agent: Audit status handling normalization. (6557fa8)
  • planner: Prevented completed task nodes from being incorrectly set to deprecated. (e0be529)
  • executor: Ensured globally unique step_id values to avoid node conflicts. (2831a96)
  • web: Fixed blank agent log and abnormal status banner display. (3d2a049)

​
Docs

  • readme: Added Tencent Cloud Hackathon badge. (6dc257a)