fix(auto-config): include all 14 tools in PERPLEXITY-MCP rules block

The auto-managed PERPLEXITY-MCP-START/END block in CLAUDE.md,
AGENTS.md, GEMINI.md, and other rules-format IDE targets listed only
10 tools. The mcp-server actually registers 14 tools in
packages/mcp-server/src/tools.ts. Missing from the rules block:

  - perplexity_export        (tools.ts:647)
  - perplexity_sync_cloud    (tools.ts:687)
  - perplexity_hydrate_cloud_entry (tools.ts:720)
  - perplexity_doctor        (tools.ts:852)

The fix is in the WRITER source -- hand-edits to the .md files would
just be re-overwritten on the next "Configure for All" run. Approach:
extracted the tool catalog into a new exported PERPLEXITY_TOOL_CATALOG
constant (still hardcoded -- adding the 4 missing entries -- but now
data instead of inline strings) and rewrote getPerplexityRulesContent
to render the bullet list from it. Also exported
getPerplexityRulesContent so tests can call it directly.

A dynamic-from-tools.ts approach was considered but rejected because
tools.ts imports the MCP SDK and heavy native deps that would pollute
the extension typecheck path. Instead, the new test parses tools.ts
with a regex (enabledTools\.has\("(perplexity_[a-z_]+)"\)) at test
time and asserts the catalog matches the registered set -- preventing
future drift without runtime coupling between the extension and
mcp-server packages.

Tests added (packages/extension/tests/auto-config.tool-catalog.test.ts,
7 assertions):
  - registered-count sanity floor (>= 14)
  - every-registered-in-catalog (the staleness check)
  - no-phantom-tools in catalog
  - no-duplicates in catalog
  - every-name-in-rendered-block
  - marker-pair wraps the block correctly
  - every-summary non-empty

Validation:
  - npm run typecheck: clean across 4 packages.
  - npm run build: clean.
  - npm run test:coverage: 119 files / 1037 pass / 2 skip / 0 fail
    (full count includes Slice 3's tests committed in 0a003f3).

NOTE: the impl agent self-reported a `git stash` baseline-check
during execution and reversed it via `stash pop`. The post-pop
worktree was independently verified by the parent before this
commit -- only the two scoped files are modified, and gates pass
cleanly.

The CLAUDE.md / AGENTS.md / GEMINI.md files in users' workspaces
will pick up the new tool entries on their next "Configure for All"
invocation. No action required by users beyond running that command.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: A.R.

packages/extension/src/auto-config/index.ts

@@ -1260,23 +1260,48 @@ export function removeTarget(target: IdeTarget): void {
   removeIdeConfig(target);
 }
 
-function getPerplexityRulesContent(): string {
+/**
+ * Hardcoded tool catalog used to render the auto-managed `PERPLEXITY-MCP-START`
+ * block in CLAUDE.md / AGENTS.md / GEMINI.md and the per-IDE rules files.
+ *
+ * Source-of-truth for the actual MCP runtime is
+ * [packages/mcp-server/src/tools.ts](../../../mcp-server/src/tools.ts) — every
+ * `if (!enabledTools || enabledTools.has("perplexity_<name>"))` branch there
+ * MUST appear in this list, otherwise the rules block downstream agents read
+ * goes stale and they call tools they think don't exist (or skip ones that do).
+ *
+ * If you add or remove a tool in `tools.ts`, update this list in the same PR.
+ * The completeness is enforced by `auto-config.tool-catalog.test.ts`, which
+ * imports the source-of-truth tool names and fails if any are missing here.
+ */
+export const PERPLEXITY_TOOL_CATALOG: ReadonlyArray<{ name: string; summary: string }> = [
+  { name: "perplexity_search", summary: "Fast web search with source citations. Use for quick factual lookups. Works with or without authentication." },
+  { name: "perplexity_reason", summary: "Step-by-step reasoning with web context. Requires Pro account." },
+  { name: "perplexity_research", summary: "Deep multi-section research reports (30-120s). Requires Pro account." },
+  { name: "perplexity_ask", summary: "Flexible queries with explicit model/mode/follow-up control." },
+  { name: "perplexity_compute", summary: "ASI/Computer mode for complex multi-step tasks. Requires Max account." },
+  { name: "perplexity_models", summary: "List available models, account tier, and rate limits." },
+  { name: "perplexity_retrieve", summary: "Poll results from pending research/compute tasks." },
+  { name: "perplexity_export", summary: "Export a saved history entry as PDF, markdown, or DOCX. Uses Perplexity's native export when available." },
+  { name: "perplexity_sync_cloud", summary: "Sync Perplexity cloud history into the local history store." },
+  { name: "perplexity_hydrate_cloud_entry", summary: "Hydrate a single cloud-backed history entry by id." },
+  { name: "perplexity_list_researches", summary: "List saved research history with status." },
+  { name: "perplexity_get_research", summary: "Fetch full content of a saved research." },
+  { name: "perplexity_login", summary: "Open browser for Perplexity authentication." },
+  { name: "perplexity_doctor", summary: "Run diagnostic checks against your Perplexity MCP install. Returns a Markdown report; pass probe:true for a live search probe." },
+];
+
+export function getPerplexityRulesContent(): string {
+  const toolLines = PERPLEXITY_TOOL_CATALOG.map(
+    ({ name, summary }) => `- **${name}** — ${summary}`,
+  );
   return [
     PERPLEXITY_RULES_SECTION_START,
     "# Perplexity MCP Server",
     "",
     "## Available Tools",
     "",
-    "- **perplexity_search** — Fast web search with source citations. Use for quick factual lookups. Works with or without authentication.",
-    "- **perplexity_reason** — Step-by-step reasoning with web context. Requires Pro account.",
-    "- **perplexity_research** — Deep multi-section research reports (30-120s). Requires Pro account.",
-    "- **perplexity_ask** — Flexible queries with explicit model/mode/follow-up control.",
-    "- **perplexity_compute** — ASI/Computer mode for complex multi-step tasks. Requires Max account.",
-    "- **perplexity_models** — List available models, account tier, and rate limits.",
-    "- **perplexity_retrieve** — Poll results from pending research/compute tasks.",
-    "- **perplexity_list_researches** — List saved research history with status.",
-    "- **perplexity_get_research** — Fetch full content of a saved research.",
-    "- **perplexity_login** — Open browser for Perplexity authentication.",
+    ...toolLines,
     "",
     "## Usage Guidelines",
     "",

packages/extension/tests/auto-config.tool-catalog.test.ts

@@ -0,0 +1,90 @@
+import { readFileSync } from "node:fs";
+import { join } from "node:path";
+import { describe, expect, it } from "vitest";
+import {
+  PERPLEXITY_RULES_SECTION_END,
+  PERPLEXITY_RULES_SECTION_START,
+} from "@perplexity-user-mcp/shared";
+import {
+  PERPLEXITY_TOOL_CATALOG,
+  getPerplexityRulesContent,
+} from "../src/auto-config/index.js";
+
+/**
+ * Source-of-truth path. The MCP server registers tools via guarded blocks like
+ *
+ *   if (!enabledTools || enabledTools.has("perplexity_<name>")) {
+ *
+ * — extracting the names with this regex matches the actual runtime registry
+ * without forcing this test to import the runtime (which pulls in the SDK and
+ * heavy native deps that aren't part of the extension package's typecheck path).
+ */
+const TOOLS_TS_PATH = join(
+  __dirname,
+  "..",
+  "..",
+  "mcp-server",
+  "src",
+  "tools.ts",
+);
+const TOOL_GUARD_RE = /enabledTools\.has\("(perplexity_[a-z_]+)"\)/g;
+
+function readRegisteredToolNames(): string[] {
+  const source = readFileSync(TOOLS_TS_PATH, "utf8");
+  const names = new Set<string>();
+  for (const match of source.matchAll(TOOL_GUARD_RE)) {
+    names.add(match[1]);
+  }
+  return [...names].sort();
+}
+
+describe("auto-config rules block tool catalog", () => {
+  const registered = readRegisteredToolNames();
+
+  it("source-of-truth has at least the audited count (sanity check)", () => {
+    // Audit at the time of writing said 14 tools registered. If this drops,
+    // the regex above probably broke; if it grows, that's expected — fix the
+    // catalog list and re-run.
+    expect(registered.length).toBeGreaterThanOrEqual(14);
+  });
+
+  it("every registered tool appears in PERPLEXITY_TOOL_CATALOG", () => {
+    const cataloged = new Set(PERPLEXITY_TOOL_CATALOG.map((t) => t.name));
+    const missing = registered.filter((name) => !cataloged.has(name));
+    expect(missing, `tools.ts registers these but the catalog omits them — update PERPLEXITY_TOOL_CATALOG: ${missing.join(", ")}`).toEqual([]);
+  });
+
+  it("catalog has no entries that aren't registered (no phantom tools)", () => {
+    const registeredSet = new Set(registered);
+    const phantom = PERPLEXITY_TOOL_CATALOG
+      .map((t) => t.name)
+      .filter((name) => !registeredSet.has(name));
+    expect(phantom, `catalog lists tools not registered in tools.ts — remove them: ${phantom.join(", ")}`).toEqual([]);
+  });
+
+  it("catalog has no duplicate tool names", () => {
+    const names = PERPLEXITY_TOOL_CATALOG.map((t) => t.name);
+    const unique = new Set(names);
+    expect(unique.size).toBe(names.length);
+  });
+
+  it("rendered rules block contains every registered tool name", () => {
+    const rendered = getPerplexityRulesContent();
+    for (const name of registered) {
+      expect(rendered, `expected rules block to mention "${name}"`).toContain(name);
+    }
+  });
+
+  it("rendered rules block is wrapped in the marker pair", () => {
+    const rendered = getPerplexityRulesContent();
+    expect(rendered.startsWith(PERPLEXITY_RULES_SECTION_START)).toBe(true);
+    expect(rendered.endsWith(PERPLEXITY_RULES_SECTION_END)).toBe(true);
+  });
+
+  it("each catalog entry has a non-empty summary", () => {
+    for (const entry of PERPLEXITY_TOOL_CATALOG) {
+      expect(entry.summary, `tool ${entry.name} missing summary`).toBeTruthy();
+      expect(entry.summary.length, `tool ${entry.name} summary too short`).toBeGreaterThan(10);
+    }
+  });
+});