#81
Feature Request: Expand AI Agent Support (Antigravity CLI, Grok-Build, GitHub Copilot Agent CLI, and custom agents)
Feature Request: Expand AI Agent Support (Antigravity CLI, Grok-Build, GitHub Copilot Agent CLI, and custom agents)
Summary
Currently, lat.md provides integrations and workflow support for several popular AI coding assistants. To make lat.md accessible to a broader ecosystem of developer tooling, we propose adding official integration guidelines, prompt templates, rule files, and CLI helper commands/configurations for additional AI coding agents—specifically Antigravity (agy) CLI, Grok-Build, GitHub Copilot Agent CLI, and generic MCP/agent extensions.
Proposed Feature / Enhancements
1. Dedicated Integration Rules & Workflows
- Antigravity (agy) CLI:
- Provide native rule/skill integration support for Antigravity agents (e.g., custom skills or
.gemini/config/rules/forlat.md). - Ensure agents running via
agyautomatically recognizelat.md/context, leveragelat search, and runlat checkprior to completing turns.
- Provide native rule/skill integration support for Antigravity agents (e.g., custom skills or
- Grok-Build:
- Add configuration templates and system prompt snippets tailored for Grok-Build pipelines to enforce
lat.mdschema validation and referential integrity checking (lat check).
- Add configuration templates and system prompt snippets tailored for Grok-Build pipelines to enforce
- GitHub Copilot Agent CLI:
- Add instructions, prompt files, or custom agents/instructions files (e.g.
.github/copilot-instructions.md) optimized for GitHub Copilot Agent CLI workflows to interact withlat.mdknowledge graphs.
- Add instructions, prompt files, or custom agents/instructions files (e.g.
2. Standardized Agent Tooling Protocol / MCP Server Support
- Provide a standardized MCP (Model Context Protocol) tool wrapper or guidelines around
lat searchandlat checkso any custom agent adhering to standard agent protocols can query thelat.mdknowledge graph seamlessly.
3. Documentation & Setup Presets
- Add a
lat init --agent <name>flag or presets inlat.mdto automatically generate agent-specific configuration/instruction files (e.g.,lat init --agent agy,lat init --agent copilot, orlat init --agent grok).
Benefits
- Broader Ecosystem Reach: Enables developers using
agyCLI,Grok-Build, orGitHub Copilot Agent CLIto maintain codebase knowledge graphs effortlessly. - Reduced Drift: Ensuring AI agents run
lat checkacross different CLI ecosystems prevents documentation rot regardless of which assistant tool developers prefer.
Additional Context / Implementation Notes
We are willing to help test or submit PRs for skill/instruction definitions and integration docs for agy CLI, Copilot Agent CLI, and other agent tools.
Activity
0No activity yet.