Merge branch 'main' into dependabot/github_actions/minor-and-patch-updates-e415389c79

rajbos · web-flow · commit 63483ef7738f · 2026-03-11T23:23:31.000+01:00
diff --git a/.github/agents/tool-names.agent.md b/.github/agents/tool-names.agent.md
@@ -0,0 +1,128 @@
+---
+description: "Add friendly display names for unknown MCP and VS Code tools detected in session logs. Handles tool name mapping in src/toolNames.json."
+name: "Tool Names - Add Missing Friendly Names"
+tools: ["execute/runInTerminal", "execute/getTerminalOutput", "search/codebase", "read/problems"]
+---
+
+# Tool Names - Add Missing Friendly Names
+
+Resolve issues that report unknown/missing tool display names by adding entries to `src/toolNames.json`. This agent understands MCP (Model Context Protocol) server naming conventions and can research tool origins.
+
+## When to Use This Agent
+
+Trigger this agent when:
+- An issue reports unknown or missing friendly tool names (title like "Add missing friendly names for tools")
+- New MCP tools appear in user session logs without display labels
+- The sync-toolnames workflow detects new upstream tools from `microsoft/vscode-copilot-chat`
+
+## Key File
+
+**`src/toolNames.json`** — The single mapping file from raw tool identifiers to human-readable display names. Every tool the extension encounters gets looked up here; missing entries show as "Unknown" in the UI.
+
+## MCP Tool Name Conventions
+
+MCP tools follow predictable naming patterns. The raw tool identifier encodes the MCP server origin and the action:
+
+### Prefix Patterns (raw identifier → display prefix)
+
+| Raw Prefix | Display Prefix | Source Repository |
+|---|---|---|
+| `mcp.io.github.git.` or `mcp_io_github_git_` | `GitHub MCP (Local):` | [github/github-mcp-server](https://github.com/github/github-mcp-server) |
+| `mcp_github_github_` | `GitHub MCP (Remote):` | [github/github-mcp-server](https://github.com/github/github-mcp-server) |
+| `mcp_github-code-s_` | `GitHub MCP (Code Scanning):` | [github/github-mcp-server](https://github.com/github/github-mcp-server) |
+| `mcp_com_microsoft_` | `GitHub MCP:` | Microsoft internal MCP server |
+| `mcp_gitkraken_` | `GitKraken MCP:` | GitKraken (no public MCP server repo; tools come from the GitKraken VS Code extension) |
+| `mcp_oraios_serena_` | `Serena:` | [oraios/serena](https://github.com/oraios/serena) |
+| `mcp_microsoft_pla_` | `Playwright MCP:` | [microsoft/playwright-mcp](https://github.com/microsoft/playwright-mcp) |
+| `mcp_io_github_ups_` | `Context7 MCP:` | [upstash/context7](https://github.com/upstash/context7) |
+
+### How to Generate a Friendly Name
+
+1. **Identify the prefix** — match the raw tool name against the prefix table above.
+2. **Extract the action** — remove the prefix to get the action part (e.g., `mcp_io_github_git_search_code` → `search_code`).
+3. **Format the action** — replace `_`, `-`, `.` with spaces, split camelCase, then Title Case each word.
+4. **Preserve acronyms** — keep these in ALL CAPS: `VSCODE`, `MCP`, `GITHUB`, `API`, `URL`, `JSON`, `HTTP`, `HTTPS`, `CLI`, `UI`, `IO`, `ID`.
+5. **Combine** — `"<Display Prefix> <Formatted Action>"`.
+
+### Examples
+
+```
+mcp_io_github_git_search_code       → "GitHub MCP (Local): Search Code"
+mcp_github_github_issue_read        → "GitHub MCP (Remote): Issue Read"
+mcp_github-code-s_list_code_scanning_alerts → "GitHub MCP (Code Scanning): List Alerts"
+mcp_gitkraken_git_status            → "GitKraken MCP: Git Status"
+mcp_oraios_serena_find_symbol       → "Serena: Find Symbol"
+mcp_microsoft_pla_browser_navigate  → "Playwright MCP: Browser Navigate"
+mcp_io_github_ups_resolve-library-id → "Context7 MCP: Resolve Library ID"
+```
+
+### Non-MCP Tools
+
+Tools without an `mcp.` or `mcp_` prefix are VS Code built-in or Copilot tools. Use plain Title Case without a server prefix:
+
+```
+run_in_terminal    → "Run In Terminal"
+copilot_readFile   → "Read File"
+setup.agent        → "Setup Agent"
+bash               → "Bash"
+grep               → "Grep"
+```
+
+Copilot-prefixed tools (`copilot_*`) typically drop the prefix in the friendly name for brevity.
+
+## Researching Unknown MCP Servers
+
+When you encounter an MCP tool with an **unfamiliar prefix** that does not match the table above:
+
+1. **Search GitHub** for the MCP server's source repository:
+   - Search for repos with topic `mcp-server` matching keywords from the prefix
+   - Check https://github.com/github/github-mcp-server for the official GitHub MCP Server tools reference
+   - Check https://github.com/punkpeye/awesome-mcp-servers for a curated community list
+   - Search `modelcontextprotocol` org repos: https://github.com/modelcontextprotocol
+2. **Inspect the server's tool definitions** — MCP servers typically expose tool names via a manifest or handler registration. Look for files like `server.json`, tool handler source files, or README documentation.
+3. **Derive the display prefix** — use the server's product/project name (e.g., "Serena", "GitKraken MCP", "Context7 MCP").
+4. **Add the new prefix pattern** to this document's table for future reference.
+
+### Well-Known MCP Server Repositories
+
+Use these repos to look up tool definitions when needed:
+
+| Server | Repository | Language | Tools File / Path |
+|---|---|---|---|
+| GitHub MCP Server | [github/github-mcp-server](https://github.com/github/github-mcp-server) | Go | `pkg/github/*.go` (each file exposes tools for a domain: issues, PRs, repos, search, actions, code scanning, etc.) |
+| Playwright MCP | [microsoft/playwright-mcp](https://github.com/microsoft/playwright-mcp) | TypeScript | Browser automation tools |
+| Serena | [oraios/serena](https://github.com/oraios/serena) | Python | Semantic code editing and symbol navigation |
+| Context7 | [upstash/context7](https://github.com/upstash/context7) | TypeScript | Library documentation retrieval |
+| Chrome DevTools MCP | [ChromeDevTools/chrome-devtools-mcp](https://github.com/ChromeDevTools/chrome-devtools-mcp) | TypeScript | Browser debugging tools |
+
+## Editing `src/toolNames.json`
+
+### Style Rules
+
+- Use leading comma style: `,"new_tool": "Friendly Name"`
+- Group related tools together (same MCP server prefix, same tool category)
+- Insert new MCP entries near existing entries with the same prefix
+- Insert new non-MCP entries alphabetically or near logically related tools
+- Never remove existing entries
+
+### Validation
+
+After editing `src/toolNames.json`:
+
+1. Run `npm run compile` to verify ESLint + build passes
+2. Ensure the JSON is valid (no trailing commas, proper quoting)
+3. Run tests with `npm run test:node` to confirm nothing is broken
+
+## Upstream Sync Reference
+
+The `sync-toolnames` workflow (`.github/workflows/sync-toolnames.yml`) automatically syncs tool IDs from `microsoft/vscode-copilot-chat` using the prompt at `.github/workflows/prompts/sync-toolnames-prompt.md`. This covers VS Code built-in and Copilot tools. MCP tool names from user sessions are **not** covered by this sync and must be added manually via issues.
+
+## Checklist
+
+- [ ] Identify all unknown tool names from the issue
+- [ ] Determine if each tool is MCP or non-MCP
+- [ ] For MCP tools, match the prefix to a known server or research the source
+- [ ] Generate friendly names following the conventions above
+- [ ] Add entries to `src/toolNames.json` in the correct location
+- [ ] Run `npm run compile` to validate
+- [ ] Run `npm run test:node` to confirm tests pass
diff --git a/.github/scripts/package-lock.json b/.github/scripts/package-lock.json
diff --git a/.github/scripts/package.json b/.github/scripts/package.json
@@ -9,6 +9,6 @@
     "scrape": "node scrape-models.js"
   },
   "dependencies": {
-    "puppeteer": "^24.38.0"
+    "puppeteer": "^24.39.0"
   }
 }
diff --git a/.github/workflows/check-toolnames.yml b/.github/workflows/check-toolnames.yml
@@ -0,0 +1,130 @@
+name: Check Toolnames in Issue
+
+on:
+  issues:
+    types: [labeled]
+
+permissions:
+  contents: read
+
+jobs:
+  # Checks each tool name in the issue body against src/toolNames.json on main
+  # and posts a comment with the status of each one.
+  check-toolnames:
+    runs-on: ubuntu-latest
+    if: github.event.label.name == 'Toolnames checkup'
+    permissions:
+      issues: write
+      contents: read
+    steps:
+      - name: Harden the runner (Audit all outbound calls)
+        uses: step-security/harden-runner@a90bcbc6539c36a85cdfeb73f7e2f433735f215b # v2.15.0
+        with:
+          egress-policy: audit
+
+      - name: Checkout main branch
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        with:
+          ref: main
+
+      - name: Check toolnames and post comment
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          ISSUE_BODY: ${{ github.event.issue.body }}
+          ISSUE_NUMBER: ${{ github.event.issue.number }}
+          REPO: ${{ github.repository }}
+          RUN_ID: ${{ github.run_id }}
+          SERVER_URL: ${{ github.server_url }}
+        run: |
+          echo "=== Checking toolnames and building comment ==="
+          
+          python3 - << 'PYEOF'
+          import json
+          import os
+          import re
+          import sys
+          
+          issue_body = os.environ.get("ISSUE_BODY", "")
+          issue_number = os.environ.get("ISSUE_NUMBER", "")
+          repo = os.environ.get("REPO", "")
+          run_id = os.environ.get("RUN_ID", "")
+          server_url = os.environ.get("SERVER_URL", "https://github.com")
+          
+          COMMENT_FILE = "/tmp/toolnames-comment.md"
+          MAX_BODY_LEN = 100_000  # guard against extremely large issue bodies
+          
+          # Truncate untrusted input to prevent resource exhaustion
+          if len(issue_body) > MAX_BODY_LEN:
+              issue_body = issue_body[:MAX_BODY_LEN]
+              print("Warning: issue body was truncated to 100 KB", file=sys.stderr)
+          
+          # Extract backtick-wrapped strings and filter to plausible tool-name shapes
+          # (no path separators or spaces; reasonable length).  This avoids matching
+          # inline-code snippets like `src/toolNames.json` or prose code examples.
+          raw_matches = re.findall(r'`([^`\n]+)`', issue_body)
+          tools = sorted(set(
+              m for m in raw_matches
+              if "/" not in m and " " not in m and len(m) <= 128
+          ))
+          print(f"Found {len(tools)} candidate tool name(s): {tools}", file=sys.stderr)
+          
+          if not tools:
+              print("No candidate tool names found in issue body. Skipping.", file=sys.stderr)
+              with open(COMMENT_FILE, "w") as f:
+                  f.write("## 🔍 Toolnames Checkup\n\n_No candidate tool names (backtick-wrapped identifiers without path separators) were found in the issue body._\n")
+          else:
+              with open("src/toolNames.json", "r") as f:
+                  known_tools = json.load(f)
+          
+              already_added = []
+              needs_adding = []
+          
+              for tool in tools:
+                  if tool in known_tools:
+                      already_added.append((tool, known_tools[tool]))
+                  else:
+                      needs_adding.append(tool)
+          
+              lines = ["## 🔍 Toolnames Checkup", ""]
+              lines.append("Checked the tool names in this issue against `src/toolNames.json` on the `main` branch:")
+              lines.append("")
+          
+              if already_added:
+                  lines.append("### ✅ Already mapped")
+                  lines.append("")
+                  lines.append("| Tool Name | Friendly Name |")
+                  lines.append("| --- | --- |")
+                  for tool, friendly in already_added:
+                      lines.append(f"| `{tool}` | {friendly} |")
+                  lines.append("")
+          
+              if needs_adding:
+                  lines.append("### ❌ Not yet in `toolNames.json`")
+                  lines.append("")
+                  lines.append("| Tool Name |")
+                  lines.append("| --- |")
+                  for tool in needs_adding:
+                      lines.append(f"| `{tool}` |")
+                  lines.append("")
+          
+              if already_added and not needs_adding:
+                  lines.append("> **All tool names in this issue are already mapped.** This issue may be a duplicate — consider closing it.")
+              elif needs_adding:
+                  lines.append(f"> **{len(needs_adding)} tool name(s) still need to be added.** Please use the `tool-names` custom agent or update `src/toolNames.json` manually.")
+          
+              lines.append("")
+              lines.append(f"_Workflow run: {server_url}/{repo}/actions/runs/{run_id}_")
+          
+              comment_body = "\n".join(lines)
+              with open(COMMENT_FILE, "w") as f:
+                  f.write(comment_body)
+              print("=== Comment preview ===", file=sys.stderr)
+              print(comment_body, file=sys.stderr)
+          PYEOF
+          
+          echo "=== Posting comment to issue #${ISSUE_NUMBER} ==="
+          gh issue comment "$ISSUE_NUMBER" \
+              --repo "$REPO" \
+              --body-file /tmp/toolnames-comment.md
+          
+          echo "✅ Comment posted successfully"
diff --git a/.github/workflows/prompts/sync-models-prompt.md b/.github/workflows/prompts/sync-models-prompt.md
@@ -15,6 +15,8 @@ Update `src/tokenEstimators.json` and `src/modelPricing.json` with missing model
    - Add it to the appropriate JSON file(s)
    - Use sensible defaults based on existing similar models
 
+> **IMPORTANT**: Only add models that are **explicitly listed** on the documentation page above. Do NOT add models based on your own knowledge of AI models, third-party sources, or speculation about what models might exist. If a model is not present on that specific documentation page, it must not be added.
+
 ## Token Estimators (`src/tokenEstimators.json`)
 
 For missing models in tokenEstimators.json:
@@ -53,7 +55,7 @@ For missing models in modelPricing.json:
 - **ONLY** update `lastUpdated` field in `src/modelPricing.json` to today's date (YYYY-MM-DD format) **if you added new models to the pricing file**
 - If no models were added, do NOT update the `lastUpdated` field
 - Do NOT modify the `sources` section unless you have specific pricing data
-- Do not make up new model names. If they are not in any of the lists in the documentation, then we do not add them to the json
+- Do NOT add models that are not explicitly listed on the documentation page — if a model is not on that page, skip it entirely
 
 ## Output Format
 
@@ -64,6 +66,7 @@ For missing models in modelPricing.json:
 
 ## Constraints
 
+- **Only add models that appear on the GitHub Copilot supported models documentation page** (`https://docs.github.com/en/copilot/reference/ai-models/supported-models`). Do NOT add models from any other source, from your own training knowledge, or that you believe might exist — only what is explicitly listed on that page.
 - Only modify `src/tokenEstimators.json` and `src/modelPricing.json`
 - Do not open a PR (the workflow will handle that)
 - Preserve all existing entries and formatting conventions
diff --git a/src/modelPricing.json b/src/modelPricing.json
@@ -2,7 +2,7 @@
   "$schema": "http://json-schema.org/draft-07/schema#",
   "description": "Model pricing data - costs per million tokens for input and output",
   "metadata": {
-    "lastUpdated": "2026-03-02",
+    "lastUpdated": "2026-03-09",
     "sources": [
       {
         "name": "OpenAI API Pricing",
@@ -115,6 +115,13 @@
       "tier": "premium",
       "multiplier": 1
     },
+    "gpt-5.4": {
+      "inputCostPerMillion": 0.00,
+      "outputCostPerMillion": 0.00,
+      "category": "GPT-5 models",
+      "tier": "premium",
+      "multiplier": 1
+    },
     "gpt-4": {
       "inputCostPerMillion": 3.0,
       "outputCostPerMillion": 12.0,
diff --git a/src/tokenEstimators.json b/src/tokenEstimators.json
@@ -20,6 +20,7 @@
     "gpt-5.2-codex": 0.25,
     "gpt-5.2-pro": 0.25,
     "gpt-5.3-codex": 0.25,
+    "gpt-5.4": 0.25,
     "gpt-4.1-nano": 0.25,
     "gemini-2.0-flash": 0.25,
     "gemini-2.0-flash-lite": 0.25,
diff --git a/src/toolNames.json b/src/toolNames.json

Original file line number	Diff line number	Diff line change
`@@ -9,6 +9,6 @@`
`9`	`9`	`"scrape": "node scrape-models.js"`
`10`	`10`	`},`
`11`	`11`	`"dependencies": {`
`12`		`- "puppeteer": "^24.38.0"`
	`12`	`+ "puppeteer": "^24.39.0"`
`13`	`13`	`}`
`14`	`14`	`}`