Skip to content

AGENTS.md

Latest commit

 

History

History
45 lines (35 loc) · 1.74 KB

AGENTS.md

File metadata and controls

45 lines (35 loc) · 1.74 KB

Agents

This agents.md is the central root file for AI agent instructions in this project. All agents and skills share and use this file alongside .ai/ directory.

Role of Agents.md

This file acts as the primary orchestrator that defines who calls which agents or skills, and delegates the tasks effectively.

Architecture

  • Root definition: agents.md (this file)
  • Subagents, AI agents, and skills exist within the .ai/ directory.

Agents

  • lead-architect: Oversees everything in the project, coordinates other AI agents, delegates workflows, and validates project progress.

Skills

  • update-agents: Triggered first on every chat/user input to check whether any instructions, patterns, or routing behaviors need to be updated in agents.md, or inside individual agent/skill files.
  • chat-log: Used on every chat/user input to prepend log entries containing timestamps, user intents, and responses to the msg-log.md table.

Global Jargon

  • menu: Means the Abby extension menu from Chrome top toolbar (extension popup panel).
  • popup: Means the Abby floating popup shown on LinkedIn pages when Abby is enabled.
  • Words written like 'search', 'apply', 'step', 'info' refer to popup tabs.

Global Rules

  • Style Compatibility: Every code update must keep UI styles readable and aligned in both light mode and dark mode.

Standard Response Format

For every response, at the final, describe the following required elements in this exact layout:

### Summary
**What you wanted:**
1. [Goal 1]
2. [Goal 2]
etc.

**What I have done:**
1. [Action 1]
2. [Action 2]
etc.

**What failed:**
- [List any failures or write "None"]

**Recommended actions:**
- [List any recommended next steps]