agent0ai/agent-zero

full scan as JSON

• Prefer Imports, Not GlobalsWhen wiring state/functions, avoid implicit access through `window.*`/`globalThis`. Use explicit ES module imports so code is determinist...

  agent0ai/agent-zero Code Style JavaScript ai-agents 2026-03-08

  copy raw open
• Centralize shared fieldsWhen multiple modules need the same state/config/schema (e.g., context variables like timezone, or prompt search paths), do not replicate...

  agent0ai/agent-zero Code Style Python ai-agents 2026-02-12

  copy raw open
• Namespace plugin APIsWhen exposing data to the client, treat shared/global values and plugin routes as explicit API surface: 1) Don’t publish global client/wi...

  agent0ai/agent-zero API Python ai-agents 2026-02-12

  copy raw open
• Endpoint Auth ValidationFor security-sensitive flows (WS/API, authenticated cloning, token/API-key auth), don’t rely on generic or heuristic validation. Apply th...

  agent0ai/agent-zero Security Python ai-agents 2026-05-07

  copy raw open
• Lock Writes, Snapshot ReadsUse the lock to protect shared mutable state during mutations, but avoid holding it for read-side convenience/field-by-field getters. **R...

  agent0ai/agent-zero Concurrency Python ai-agents 2026-01-10

  copy raw open
• Single Source Error LogicWhen handling failures or optional data, make the decision exactly once and reuse the result everywhere. Apply two concrete rules: 1) **N...

  agent0ai/agent-zero Error Handling Python ai-agents 2026-04-27

  copy raw open
• Unified extension API contractsWhen building WebUI plugin/extension systems, treat the backend as the source of truth for (1) resolving extension contents and (2) provi...

  agent0ai/agent-zero API JavaScript ai-agents 2026-02-25

  copy raw open
• AI Boundary CompatibilityWhen building AI/LLM-related tooling (tool-calls, background execution, inference/image generation), enforce compatibility at the boundar...

  agent0ai/agent-zero AI Python ai-agents 2025-12-29

  copy raw open
• Config source precedenceWhen configuration is moved/renamed or when settings are derived from environment variables, ensure (1) the dotenv loader points to the c...

  agent0ai/agent-zero Configurations Python ai-agents 2025-12-11

  copy raw open
• Sanitize Model OutputWhen handling untrusted model/text output that will be serialized into JSON (or rendered in a UI), normalize it before output generation—...

  agent0ai/agent-zero Security Other ai-agents 2025-04-08

  copy raw open
• Sanitize untrusted tagsTreat any tag/markup coming from user input, LLM output, or external commands as untrusted. Validate and sanitize it at the earliest boun...

  agent0ai/agent-zero Security Shell ai-agents 2025-04-08

  copy raw open