Skip to content
Components

Components

Existing components to architect and optimize agents’ behavioral work.

Explore

(Custom) Agent
Sessions handling review, investigation, build, or other varied tasks all within …
Command / Prompt
Commands (or prompts) invoke custom instructions on demand through a slash …
Hook
Hooks are deterministic configurations to execute a command, a prompt, a webhook …
Instruction / Rule
Instructions are known to be the global rules an agent must respect when acting, …
LSP
Language Server Protocol (LSP) has been there way before AI arrived. This is …
MCP
When in need of documentation regarding tools, language features, interactions …
Monitor
Long-running or external processes (e.g. a deployment rolling out, a log file …
Skill
Within their tasks, agents may need to handle specific smaller actions: …

Mental Map 

    flowchart TD
    Agent{"Agent"}
    Component["Command / Prompt, LSP, MCP, Skill"]
    GlobalInstructions["Global instructions"]
    Instruction["Instructions specific to files"]
    Monitor[["Monitor"]]
    SubAgent{"Agent"}
    ToolUse["Bash, Write, Read, etc."]
    User(["User"])

    User ---|prompts| Hook1[["Hook"]] --> Agent
    Agent ---|uses| Hook2[["Hook"]] --> ToolUse
    Agent -->|uses / requests| Component
    Agent -->|reads| Instruction
    Agent ---|spawns| Hook3[["Hook"]] --> SubAgent

    GlobalInstructions -.->|shapes global behaviors| Agent
    Monitor -.->|notifies| Agent

    NoteSubAgent["Behaves like Agent: uses, reads, shaped/notified the same way"]
    SubAgent -.- NoteSubAgent

    classDef note fill:none,stroke-dasharray: 3 3
    class NoteSubAgent note