agent-gauntlet 0.1.5 → 0.1.7

package/README.md

@@ -18,22 +18,22 @@ Agent Gauntlet is designed to be "tool-agnostic" by leveraging the AI CLI tools
 
 ### Requirements
 
-- **Bun** or **Node.js** (runtime)
+- **Bun** (Required runtime, v1.0.0+)
 - **git** (change detection and diffs)
 - For review gates: one or more supported AI CLIs installed (`gemini`, `codex`, `claude`). For the full list of tools and how they are used, see [CLI Invocation Details](docs/cli-invocation-details.md)
 
 ### Installation
 
-You can install `agent-gauntlet` globally using `npm` or `bun`:
+You can install `agent-gauntlet` globally using `npm` or `bun` (Bun must be installed on the system in both cases):
 
-**Using npm:**
+**Using Bun (Recommended):**
 ```bash
-npm install -g agent-gauntlet
+bun add -g agent-gauntlet
 ```
 
-**Using Bun (Recommended):**
+**Using npm:**
 ```bash
-bun add -g agent-gauntlet
+npm install -g agent-gauntlet
 ```
 
 ### Quick start

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agent-gauntlet",
-  "version": "0.1.5",
+  "version": "0.1.7",
   "description": "A CLI tool for testing AI coding agents",
   "license": "Apache-2.0",
   "author": "Paul Caplan",

package/src/commands/init.ts

@@ -13,12 +13,12 @@ allowed-tools: Bash
 # /gauntlet
 Execute the autonomous verification suite.
 
-1. Run \`npx agent-gauntlet run\` (or \`./bin/agent-gauntlet run\` if developing locally).
+1. Run \`agent-gauntlet run\`.
 2. If it fails, read the log files in \`.gauntlet_logs/\` to understand exactly what went wrong.
 3. Fix any code or logic errors found by the tools or AI reviewers, prioritizing higher-priority violations (critical > high > medium > low).
 4. If you disagree with AI reviewer feedback, briefly explain your reasoning in the code comments rather than ignoring it silently.
 5. Do NOT commit your changes yet—keep them uncommitted so the rerun command can review them.
-6. Run \`npx agent-gauntlet rerun\` (or \`./bin/agent-gauntlet rerun\` if developing locally) to verify your fixes. The rerun command reviews only uncommitted changes and uses previous failures as context.
+6. Run \`agent-gauntlet rerun\` to verify your fixes. The rerun command reviews only uncommitted changes and uses previous failures as context.
 7. Repeat steps 2-6 until one of the following termination conditions is met:
    - All gates pass
    - You disagree with remaining failures (ask the human how to proceed)