Skip to content

docs(profile): collapse the AI-agents section to one paragraph + guide link #44

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
rafael5 merged 1 commit into from
May 11, 2026
+1 −22
Merged

docs(profile): collapse the AI-agents section to one paragraph + guide link #44

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 1 addition & 22 deletions profile/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,28 +23,7 @@ maintaining modern (non-VistA) M code.

## For AI agents and automated tooling

**m-dev-tools is fully AI-enabled with an MCP endpoint.** The fastest path is to point any MCP-capable agent (Claude Code, Codex, Continue, …) at [`m-dev-tools-mcp`](https://github.com/m-dev-tools/m-dev-tools-mcp); drop this into your project's `.mcp.json`:

```json
{
"mcpServers": {
"m-dev-tools": {
"command": "uvx",
"args": ["--from", "git+https://github.com/m-dev-tools/m-dev-tools-mcp@v0.1.0", "m-dev-tools-mcp"]
}
}
}
```

The server exposes three tools — `route_intent("parse JSON in M") → ["module:m-stdlib#STDJSON"]`, `describe(typed_id) → pointer-blob`, `verify(repo) → verification_commands` (lists, doesn't execute) — that walk the catalog for you and ground every claim in a manifest URL the agent actually fetched. No fabricated APIs.

**Full walkthrough:** [`docs/ai-discoverability/ai-users-guide.md`](../docs/ai-discoverability/ai-users-guide.md) — install paths, example sessions, troubleshooting, where to file bugs.

**Direct catalog access** (no MCP) is still supported — start at [`tools.json`](tools.json) (machine-readable catalog of every repo + every `*_url` pointer + `consumed_by` graph) and [`task_index.json`](task_index.json) (hand-curated intent → typed-ID routing). The 8-step discovery handshake in [`AI-discoverability-architecture.md`](../docs/ai-discoverability/AI-discoverability-architecture.md) spells out the canonical sequence. Four continuous-enforcement gates (freshness / link-check / license-reconcile / schema-version) keep the catalog honest — every PR + a weekly cron firing. See [`docs/ai-discoverability/`](../docs/ai-discoverability/) for the full framework.

Manifest pointers for symbol-level lookups:
- **m-stdlib API** → [`dist/stdlib-manifest.json`](https://raw.githubusercontent.com/m-dev-tools/m-stdlib/main/dist/stdlib-manifest.json)
- **M language surface** → [`integrated/grammar-surface.json`](https://raw.githubusercontent.com/m-dev-tools/m-standard/main/integrated/grammar-surface.json)
m-dev-tools ships an MCP server ([`m-dev-tools-mcp`](https://pypi.org/project/m-dev-tools-mcp/), also listed on the [official MCP registry](https://registry.modelcontextprotocol.io/) as `io.github.m-dev-tools/m-dev-tools-mcp`) that exposes `route_intent` / `describe` / `verify` over the catalog so any MCP-capable agent (Claude Code, Codex, Continue, …) can resolve plain-English M-tooling intent without guessing. See the [**AI users guide**](../docs/ai-discoverability/ai-users-guide.md) for install paths (PyPI, `.mcp.json`, registry-driven), example sessions, and the no-MCP fallback that walks `tools.json` directly.

## Repositories

Expand Down
Loading