Merge pull request #106 from microsoft/ozzafar/reduce_vsce_package_files

make skill registration harness generic

Author: ozzafar

.gitignore

@@ -420,4 +420,5 @@ FodyWeavers.xsd
 
 node_modules/*
 out/*
+dist/*
 .vscode-test/*

.vscodeignore

@@ -1,12 +1,18 @@
 .vscode/**
 .vscode-test/**
+.github/**
 src/**
+out/**
+node_modules/**
+docs/**
+debugmcp-ad/**
+.pytest_cache/**
 .gitignore
 .yarnrc
+esbuild.js
 vsc-extension-quickstart.md
 **/tsconfig.json
 **/eslint.config.mjs
 **/*.map
 **/*.ts
 **/.vscode-test.*
-debugmcp-ad/**

README.md

@@ -4,7 +4,7 @@ Let AI agents debug your code inside VS Code - set breakpoints, step through exe
 
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 [![VS Code](https://img.shields.io/badge/VS%20Code-1.104.0+-blue.svg)](https://code.visualstudio.com/)
-[![Version](https://img.shields.io/badge/version-2.2.0-green.svg)](https://github.com/microsoft/DebugMCP)
+[![Version](https://img.shields.io/badge/version-2.2.1-green.svg)](https://github.com/microsoft/DebugMCP)
 [![VS Marketplace](https://img.shields.io/badge/VS%20Marketplace-Install-blue.svg)](https://marketplace.visualstudio.com/items?itemName=ozzafar.debugmcpextension)
 
 > ⭐ **If you find DebugMCP useful, please [star the repo on GitHub](https://github.com/microsoft/DebugMCP)!** It helps others discover the project and motivates continued development.
@@ -17,11 +17,12 @@ Let AI agents debug your code inside VS Code - set breakpoints, step through exe
   <img src="assets/DebugMCP.gif" width="800">
 </p>
 
-## ✨ What's New in 2.2.0
+## ✨ What's New
 
-- **`/debug-live` Agent Skill** — DebugMCP now ships a companion [Agent Skill](./skills/debug-live/SKILL.md) that is auto-installed into each configured harness's personal skills directory (e.g. `~/.copilot/skills/debug-live/`). Invoke it with `/debug-live` in supporting agents to load the systematic debugging workflow and trigger DebugMCP tools with the right context.
-- **Robust debugging via the VS Code Testing API** — `start_debugging` with a `testName` now uses the VS Code Testing API to discover and launch the test, replacing the previous best-effort path. This works reliably across language test runners that integrate with the Testing API (pytest, Jest/Vitest, Java, .NET, Go, etc.) and produces consistent breakpoint hits inside individual test cases.
-- **Concurrent debugging** - supports multiple concurrent debug sessions, allowing effective parallel agentic debugging.
+### 2.2
+- **Cross-agent `debug-live` skill install** — the systematic debugging workflow ships as an [Agent Skill](https://agentskills.io) and is now installed into the **standard skills directories** — `~/.agents/skills/` (the cross-agent location honored by skills-compatible harnesses, including VS Code agent mode) and `~/.copilot/skills/` when present — so it's discoverable everywhere instead of being copied next to each agent's config where nothing scans it (fixes [#105](https://github.com/microsoft/DebugMCP/issues/105), where VS Code never loaded the skill). The server also advertises MCP `instructions` and the `start_debugging` tool points at the skill for the full workflow.
+- **Pause running programs** — new `pause_execution` tool interrupts a freely-running program and stops at its current location, even with no breakpoint set (great for busy loops and embedded/bare-metal targets), so you can then inspect state or step from there.
+- **Robust debugging via the VS Code Testing API** — `start_debugging` with a `testName` uses the VS Code Testing API to discover and launch the test, producing consistent breakpoint hits inside individual test cases across language test runners (pytest, Jest/Vitest, Java, .NET, Go, etc.).
 
 ## 🚀 Quick Install
 
@@ -57,6 +58,7 @@ DebugMCP is an MCP server that gives AI coding agents full control over the VS C
 | **step_into** | Step into function calls | None |
 | **step_out** | Step out of the current function | None |
 | **continue_execution** | Continue until next breakpoint | None |
+| **pause_execution** | Interrupt a freely-running program and stop at its current location (no breakpoint needed) | None |
 | **restart_debugging** | Restart the current debug session | None |
 | **add_breakpoint** | Add a breakpoint at a specific line (optionally conditional) | `fileFullPath` (required)<br>`lineContent` (required)<br>`condition` (optional) |
 | **remove_breakpoint** | Remove a breakpoint from a specific line | `fileFullPath` (required)<br>`line` (required) |
@@ -65,11 +67,13 @@ DebugMCP is an MCP server that gives AI coding agents full control over the VS C
 | **get_variables_values** | Get variables and their values at current execution point | `scope` (optional: 'local', 'global', 'all') |
 | **evaluate_expression** | Evaluate an expression in debug context | `expression` (required) |
 
-> **Note:** The MCP server intentionally exposes **tools only** — no procedural
-> instructions, no documentation resources. Workflow guidance (when to debug, how to
-> structure a root-cause investigation, language-specific quirks) lives in the companion
-> [DebugMCP Agent Skill](./skills/debug-live/SKILL.md) so it can be loaded into an
-> agent's prompt context independently of the MCP capability surface.
+> **Note:** The MCP server exposes **tools** for debugger actions, while the procedural
+> workflow guidance (when to debug, how to structure a root-cause investigation,
+> language-specific quirks) lives in the companion [Agent Skill](./skills/debug-live/SKILL.md).
+> Tool descriptions stay terse and behavioral; the extension installs the `debug-live` skill
+> into the standard skills directories (`~/.agents/skills/`, plus `~/.copilot/skills/` when
+> present) so skills-compatible harnesses load the full workflow on demand. The server also
+> advertises MCP `instructions` pointing agents at it before debugging.
 
 ### 🎯 Debugging Best Practices
 

docs/architecture/agentConfigurationManager.md

@@ -73,11 +73,22 @@ url = "http://localhost:3001/mcp"
 
 Uses VS Code's `globalState` to track whether the onboarding popup has been shown, preventing repeated prompts on every activation.
 
+### Skill delivery — standard skills directories
+
+The `debug-live` Agent Skill is installed into the **standard personal skills directories** defined by the Agent Skills open standard (agentskills.io), rather than being copied next to each agent's config file:
+- **`~/.agents/skills/debug-live/`** — the cross-agent location honored by skills-compatible harnesses, including VS Code agent mode and Copilot CLI. Always installed.
+- **`~/.copilot/skills/debug-live/`** — Copilot's own skills path; also installed when a Copilot home directory (`~/.copilot`, or `$COPILOT_HOME`) exists.
+
+`installDebugMCPSkill()` copies the one bundled source (`skills/debug-live/SKILL.md`) into each target with `force: true` (idempotent refresh) and removes stale legacy copies (`debug`, `really-debug`). It is agent-independent — a single shared install covers every skills-compatible harness.
+
+This fixes issue #105: earlier builds copied the skill next to each agent's config (e.g. `Code/User/skills/` for VS Code Copilot), a directory no harness scans, so the skill never loaded. Installing to `~/.agents/skills/` — which VS Code agent mode does scan — makes it discoverable.
+
 ## Key Code Locations
 
 - Class definition: `src/utils/agentConfigurationManager.ts`
 - Agent definitions: `getSupportedAgents()`
 - Config writing: `addDebugMCPToAgent()`
+- Skill install: `installDebugMCPSkill()` / `getSkillInstallTargets()` / `ensureSkillRegistered()`
 - Codex TOML upsert: `upsertCodexDebugMCPConfig()`
 - Path detection: `getConfigBasePath()`
 - Popup logic: `shouldShowPopup()`, `showAgentSelectionPopup()`
@@ -91,6 +102,8 @@ Uses VS Code's `globalState` to track whether the onboarding popup has been show
 5. Show success message with option to open config file
 6. Mark popup as shown
 
+The bundled `debug-live` skill is installed into the standard skills directories (`~/.agents/skills/`, plus `~/.copilot/skills/` when present) during step 4, so every skills-compatible harness discovers it from one shared location.
+
 ## Commands
 
 - `debugmcp.showAgentSelectionPopup`: Manually trigger agent setup

docs/architecture/debugMCPServer.md

@@ -54,17 +54,26 @@ extension. To avoid debugging the wrong workspace when several windows are open:
 `DebugMCPServer` builds one handler **per MCP session** via a handler factory, which is
 what lets concurrent agent sessions drive debuggers in different repos simultaneously.
 
-### Tools only — no resources, no instructions
+### Tools & debug-live skill
 
-`DebugMCPServer` exposes **tools only**. Procedural workflow guidance (when to debug,
-how to structure a root-cause investigation, language-specific quirks) lives in the
-companion Agent Skill at `skills/debug-live/SKILL.md`, not in tool descriptions or MCP
-resources. This separation matches modern agent ecosystems where MCP servers provide
-*capabilities* and skills provide *procedural knowledge* an agent loads as context.
+`DebugMCPServer` exposes **tools** for debugger capabilities. Detailed procedural guidance
+(when to debug, how to structure a root-cause investigation, language-specific quirks) lives
+in the companion Agent Skill at `skills/debug-live/SKILL.md`, which `AgentConfigurationManager`
+installs into the standard personal skills directories (`~/.agents/skills/`, and
+`~/.copilot/skills/` when present) so skills-compatible harnesses load it on demand. This
+separation matches modern agent ecosystems where MCP servers provide *capabilities* and
+skills provide *procedural knowledge* an agent loads as context.
 
 Tool descriptions are intentionally terse and behavioral — they describe *what* the
 tool does, not *when* or *how* to use it in a multi-step workflow.
 
+### Ensuring the workflow is loaded
+
+The server `instructions` (passed to the `McpServer` constructor and returned to the client
+at `initialize`) and the `start_debugging` tool description both point agents at the
+`debug-live` skill for the full step-through workflow, so the pointer is visible even before
+the skill activates.
+
 ### Streamable HTTP Transport
 
 Uses stateless HTTP POST requests for MCP communication. The express server exposes:
@@ -96,7 +105,7 @@ error wins:
 - Per-window control server: `src/controlServer.ts`
 - Cross-window routing handler: `src/routingDebuggingHandler.ts`
 - Shared window registry: `src/utils/workspaceRegistry.ts`
-- Agent Skill (companion, not part of the MCP surface): `skills/debug-live/SKILL.md`
+- Agent Skill (procedural workflow): `skills/debug-live/SKILL.md`
 
 ## Exposed Tools
 
@@ -106,6 +115,7 @@ error wins:
 | `stop_debugging` | Stop current session |
 | `step_over/into/out` | Stepping commands |
 | `continue_execution` | Continue to next breakpoint |
+| `pause_execution` | Interrupt a running program (no breakpoint needed) |
 | `restart_debugging` | Restart session |
 | `add/remove_breakpoint` | Breakpoint management |
 | `clear_all_breakpoints` | Remove all breakpoints |

docs/architecture/debuggingExecutor.md

@@ -45,6 +45,7 @@ Stepping and control operations use VS Code's command system:
 - `workbench.action.debug.stepInto`
 - `workbench.action.debug.stepOut`
 - `workbench.action.debug.continue`
+- `workbench.action.debug.pause`
 - `workbench.action.debug.restart`
 
 ### DAP Custom Requests

esbuild.js

@@ -0,0 +1,57 @@
+// Copyright (c) Microsoft Corporation.
+
+const esbuild = require('esbuild');
+
+const production = process.argv.includes('--production');
+const watch = process.argv.includes('--watch');
+
+/**
+ * Bundles the extension into a single CommonJS file so the published VSIX ships
+ * one JS file instead of the entire production node_modules tree. `vscode` is
+ * kept external because it is provided by the host at runtime.
+ * @type {import('esbuild').Plugin}
+ */
+const esbuildProblemMatcherPlugin = {
+	name: 'esbuild-problem-matcher',
+	setup(build) {
+		build.onStart(() => {
+			console.log('[esbuild] build started');
+		});
+		build.onEnd((result) => {
+			result.errors.forEach(({ text, location }) => {
+				console.error(`✘ [ERROR] ${text}`);
+				if (location) {
+					console.error(`    ${location.file}:${location.line}:${location.column}`);
+				}
+			});
+			console.log('[esbuild] build finished');
+		});
+	},
+};
+
+async function main() {
+	const ctx = await esbuild.context({
+		entryPoints: ['src/extension.ts'],
+		bundle: true,
+		format: 'cjs',
+		minify: production,
+		sourcemap: !production,
+		sourcesContent: false,
+		platform: 'node',
+		outfile: 'dist/extension.js',
+		external: ['vscode'],
+		logLevel: 'silent',
+		plugins: [esbuildProblemMatcherPlugin],
+	});
+	if (watch) {
+		await ctx.watch();
+	} else {
+		await ctx.rebuild();
+		await ctx.dispose();
+	}
+}
+
+main().catch((e) => {
+	console.error(e);
+	process.exit(1);
+});