Project Files

.github

workflows

ci.yml

scripts

sync-version.mjs

skills

academic.md

market.md

policy.md

src

config

keys.ts

local

gatekeeper.ts

pdf-batch.ts

search.ts

store.ts

memory

derive.ts

persist.ts

retrieve.ts

store.ts

planning

dimensions.ts

planner.ts

schemas.ts

skills.ts

report

builder.ts

scoring

authority.ts

critic.ts

evidence.ts

fastCriticOptimizer.ts

heuristics.ts

skills

loader.ts

select.ts

swarm

blackboard.ts

health.ts

learning.ts

logger.ts

orchestrator.ts

pluginBridgeAndPacer.ts

singleflightAndSnippetGate.ts

uncertainty-miner.ts

visited-cache.ts

worker.ts

synthesis

ai.ts

verify.ts

utils

llm.ts

qwen-parser.ts

qwenJsonParser.ts

qwenOptimizer.ts

text.ts

timeBudget.ts

tokens.ts

topicQuery.ts

api-keys.json

config.ts

constants.ts

index.ts

polyfills.ts

researcher.ts

toolsProvider.ts

types.ts

.gitignore

AGENTS.md

CHANGELOG-v5.12.2.md

CHANGELOG.md

imp.md

implement.md

LICENSE

manifest.json

package.json

README.md

tsconfig.json

vitest.config.ts

tests / visited-cache.test.ts

import { afterEach, describe, expect, it, vi } from "vitest";
import fs from "node:fs";
import os from "node:os";
import path from "node:path";
import { VisitedPageCache } from "../src/swarm/visited-cache";
import { CrawledSource } from "../src/types";

const roots: string[] = [];
function cachePath(): string {
  const dir = fs.mkdtempSync(path.join(os.tmpdir(), "swarm-cache-test-"));
  roots.push(dir);
  return path.join(dir, "cache.json");
}
afterEach(() => { vi.useRealTimers(); for (const dir of roots.splice(0)) fs.rmSync(dir, { recursive: true, force: true }); });

function page(url: string, text: string): CrawledSource {
  return { url, finalUrl: url, title: "Renewable energy research", description: text, text, wordCount: 100, published: null, outlinks: [], sourceQuery: "energy", workerRole: "academic", workerLabel: "worker", domainScore: 80, freshnessScore: 50, tier: "academic", relevanceScore: 0.8, origin: "web" };
}

describe("visited page cache", () => {
  it("saves and retrieves relevant pages without rediscovering their URLs", () => {
    const file = cachePath();
    const cache = new VisitedPageCache(30, file);
    cache.markVisited(page("https://example.org/a", "Photovoltaic energy storage research"));
    cache.flush();
    const restored = new VisitedPageCache(30, file);
    expect(restored.searchRecent("photovoltaic energy storage", 3)).toHaveLength(1);
    expect(restored.searchRecent("marine biodiversity", 3)).toHaveLength(0);
  });

  it("does not reuse pages after the retention period", () => {
    vi.useFakeTimers();
    const cache = new VisitedPageCache(1, cachePath());
    cache.markVisited(page("https://example.org/a", "Photovoltaic energy storage research"));
    vi.advanceTimersByTime(2 * 86400000);
    expect(cache.searchRecent("photovoltaic energy storage", 3)).toEqual([]);
  });

  it("does not blacklist a URL for being irrelevant to a different query", () => {
    const cache = new VisitedPageCache(30, cachePath());
    cache.markRejected("https://example.org/a", "very-off-topic");
    expect(cache.isRejected("https://example.org/a")).toBe(false);
    cache.markRejected("https://example.org/a", "HTTP 403");
    expect(cache.isRejected("https://example.org/a")).toBe(true);
  });
});

import { afterEach, describe, expect, it, vi } from "vitest";
import fs from "node:fs";
import os from "node:os";
import path from "node:path";
import { VisitedPageCache } from "../src/swarm/visited-cache";
import { CrawledSource } from "../src/types";

const roots: string[] = [];
function cachePath(): string {
  const dir = fs.mkdtempSync(path.join(os.tmpdir(), "swarm-cache-test-"));
  roots.push(dir);
  return path.join(dir, "cache.json");
}
afterEach(() => { vi.useRealTimers(); for (const dir of roots.splice(0)) fs.rmSync(dir, { recursive: true, force: true }); });

function page(url: string, text: string): CrawledSource {
  return { url, finalUrl: url, title: "Renewable energy research", description: text, text, wordCount: 100, published: null, outlinks: [], sourceQuery: "energy", workerRole: "academic", workerLabel: "worker", domainScore: 80, freshnessScore: 50, tier: "academic", relevanceScore: 0.8, origin: "web" };
}

describe("visited page cache", () => {
  it("saves and retrieves relevant pages without rediscovering their URLs", () => {
    const file = cachePath();
    const cache = new VisitedPageCache(30, file);
    cache.markVisited(page("https://example.org/a", "Photovoltaic energy storage research"));
    cache.flush();
    const restored = new VisitedPageCache(30, file);
    expect(restored.searchRecent("photovoltaic energy storage", 3)).toHaveLength(1);
    expect(restored.searchRecent("marine biodiversity", 3)).toHaveLength(0);
  });

  it("does not reuse pages after the retention period", () => {
    vi.useFakeTimers();
    const cache = new VisitedPageCache(1, cachePath());
    cache.markVisited(page("https://example.org/a", "Photovoltaic energy storage research"));
    vi.advanceTimersByTime(2 * 86400000);
    expect(cache.searchRecent("photovoltaic energy storage", 3)).toEqual([]);
  });

  it("does not blacklist a URL for being irrelevant to a different query", () => {
    const cache = new VisitedPageCache(30, cachePath());
    cache.markRejected("https://example.org/a", "very-off-topic");
    expect(cache.isRejected("https://example.org/a")).toBe(false);
    cache.markRejected("https://example.org/a", "HTTP 403");
    expect(cache.isRejected("https://example.org/a")).toBe(true);
  });
});

deep-swarm-research-plugin-with-time-limit

deep-swarm-research-plugin-with-time-limit