servicetitan/arcade-eval
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: cross-tool entry points (AGENTS.md, Cursor rule, mcp.json example)

Browse Source
This commit is contained in:
iztaylor
2026-06-18 10:10:55 -04:00
parent 593e1e63b6
commit b776157c95
3 changed files with 52 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.cursor/rules/arcade-eval.mdc
+19
View File
@@ -0,0 +1,19 @@
---
description: Arcade eval lane bootstrap — read order, live-state check, lane ownership
alwaysApply: true
---
You are working in the arcade-eval repo (Arcade.dev MCP-gateway evaluation).
Before working: read README.md → STATUS.md → LIVE-POC.md → GROUND-RULES.md (in that order),
then run the live-state check described in GROUND-RULES.md.
Working rules:
- Write only inside your own categories/catN-*/ subtree and your own STATUS.md section.
- Shared files (config/targets.yaml, lib/, top-level docs) are append-mostly; `git pull --rebase` before push.
- Credentials live only in .env (git-ignored) — never commit them.
- Never edit the criteria Google Doc from a session; compose criteria-section-N.md locally, the human pastes. Criterion wording is verbatim.
Connecting Cursor as an MCP client: copy .cursor/mcp.json.example to .cursor/mcp.json
(git-ignored) and fill the gateway slug + headers. Cursor is itself a required cat-1 client
to verify (criteria: "Claude Code, Cursor, LangGraph, internal agent frameworks").