lexigram-ai-skills

AI skills and tools for the Lexigram Framework — registry, executor, builtin tools, discovery

Overview

Composable, registry-based skill execution for the Lexigram AI framework. Define skills as classes or decorated functions, execute them with retry, caching, permission enforcement, and timeout — and compose them into chains, pipelines, parallel fans, and content-routers. Zero-config usage starts with sensible defaults.

Full documentation: docs.lexigram.dev

Install

uv add lexigram-ai-skills

Quick Start

from lexigram import Application
from lexigram.di.module import Module, module

from lexigram.ai.skills import SkillsModule
from lexigram.ai.skills.config import SkillsConfig


@module(
    imports=[
        SkillsModule.configure(
            SkillsConfig(
                enable_builtin=True,
                builtin_skills=["current_datetime", "math_calculate"],
                cache_enabled=True,
                enforce_permissions=False,
            )
        )
    ]
)
class AppModule(Module):
    pass


async with Application.boot(modules=[AppModule]) as app:
    # use app.container to resolve services
    ...

Configuration

Zero-config usage: Call SkillsModule.configure() with no arguments to use defaults.

Option 1 — YAML file

# application.yaml
ai_skills:
  default_timeout_seconds: 30.0
  cache_enabled: true
  enforce_permissions: true
  enable_builtin: true
  builtin_skills:
    - "current_datetime"
    - "math_calculate"

Option 2 — Profiles + Environment Variables (recommended)

export LEX_AI_SKILLS__ENFORCE_PERMISSIONS=true
# Environment variables for each field

Option 3 — Python

from lexigram.ai.skills.config import SkillsConfig
from lexigram.ai.skills import SkillsModule

config = SkillsConfig(
    default_timeout_seconds=30.0,
    cache_enabled=True,
    enforce_permissions=True,
    enable_builtin=True,
    builtin_skills=["current_datetime", "math_calculate", "text_summarize"],
)
SkillsModule.configure(config)

Config reference

Module Factory Methods

Key Features

• Class-based skills: AbstractSkill base class with SkillDefinition
• Function-based skills: @skill decorator for registering functions
• Skill executor: Retry, caching, permissions, and timeout enforcement
• Built-in skills: DateTime, Math, HTTPRequest, WebSearch, FileRead, FileWrite, DatabaseQuery, CodeExecute
• Composition: SkillChain, ParallelSkills, SkillPipeline, SkillRouter
• MCP bridge: MCPSkillBridge imports MCP tools as skills and exports skills as MCP tools
• Registry: Named lookup and registration via SkillRegistry

Testing

async with Application.boot(modules=[SkillsModule.stub()]) as app:
    # your test code
    ...

Key Source Files