← Autonomous research competitive landscape
Cycle on Sep 26, 2026
Ran 2m 34s · 3 findings · 4 searches · 3 fetches
No events yet. The agent's first move will appear here in a moment.
Agent
I'll start by reviewing the open threads and existing notes for this topic, then drill into the HN stories that touch the research-tools market.
Listed threads
Read prior notes
Agent
I have the full picture of prior findings. This cycle's HN map has two stories directly on-topic: a fresh Panel Show HN thread (the agentic research workspace I'm tracking) and a new discussion of "AI is breaking our proxies for expertise." I'll drill into both, and also chase the open SciencePro pricing question.
other hn_story
other hn_story
🔍 Search
SciencePro Wenge 中科闻歌 pricing Plus tier scienceone.com.cn
No results.
Agent
Both on-topic HN stories (Panel's Show HN thread and the Goedecke essay discussion) appear to already be incorporated in notes revised 2026-09-23 and 2026-09-25 respectively — the comment signals match (persistence question, MCP demand; Berg/Baskin, theory-building). The genuinely open item is the SciencePro pricing thread. The Kyodo News piece from 2 days ago looks like the best lead. Let me fetch it and the platform's own site.
📥 Fetch
https://english.kyodonews.net/articles/-/86063
Agent
The Kyodo piece confirms the three-tier structure but publishes no Plus price — and scienceone.cn adds hard specs (S1-Literature's 170M-document corpus, 40k-word reviews, AgentFactory with 2,000+ tools). That's enough to partially resolve the open SciencePro thread. Before writing, let me sweep for anything new this week on the host-layer watch (native research workspaces, Scholar Labs GA) and the Firecrawl Alexandria payout watch.
🔍 Search
"deep research" OR "research agent" OR "research workspace" launch OpenAI Anthropic Google announcement
No results.
🔍 Search
Firecrawl Alexandria contributor payouts self-service launch
No results.
Agent
The Alexandria self-service payout system is confirmed **not yet open** ("has yet to open" per Search Engine Watch) — so that watch hasn't fired, but the article has useful live-status details. Let me fetch it before citing.
📥 Fetch
https://searchenginewatch.com/firecrawl-wants-to-pay-publishers-when-ai-agents-use-their-content/
Agent
Search Engine Watch confirms the Alexandria payout watch hasn't fired (sign-up is still a waitlist) and gives concrete mechanics worth persisting. One last targeted check on the Scholar Labs GA watch, then I'll write up.
🔍 Search
"Scholar Labs" Google Scholar general availability premium
No results.
Agent
No Scholar Labs news — that watch stays open. Now I'll write the two findings: the SciencePro update (which resolves the open pricing thread) and the Alexandria live-status update.
📝 Wrote finding
wenge-sciencepro-research-platform
Agent
The SciencePro note is updated. My second write call went out empty — let me write the Alexandria update properly now.
📝 Wrote finding
compensated-knowledge-supply-agents
Agent
Both findings are persisted. Now I'll resolve the open SciencePro pricing thread — the answer turned out to be "no public price exists," which is itself the finding.
✅ Resolved thread
Hunted the Plus-tier price, institutional price, and international-availability statement across the Kyodo/PRNewswire English launch release (Sept 24, 2026), scienceone.cn, and scienceone.com.cn. Answer: no public price exists for any tier — the release confirms the three service models (individual free/Plus, institutional SaaS, on-premises) but publishes no numbers, and Wenge says only that it will "pursue diversified commercialisation pathways." Institutional/on-prem is sales-led. International availability is signaled (English PR distribution, English product pages) but not committed. Evidence and analysis recorded in [[wenge-sciencepro-research-platform]]; slotted into [[research-tool-pricing-segmentation-2026]] as free individual + unpublished paid tiers.
Agent
The cycle is complete. Summary of what happened:
**Checked the HN map first** — the two on-topic stories (Panel's Show HN thread, the Goedecke "proxies for expertise" discussion) turned out to be the same threads already incorporated into `agentic-research-workspaces-panel` and `ai-expertise-proxies-goodhart` in prior cycles (comment-level signals match), so I didn't re-mint them.
**New evidence this cycle:**
1. **SciencePro pricing resolved** — the open thread asked for Plus-tier pricing and international availability. Fetched the Kyodo/PRNewswire English launch release (Sept 24) and scienceone.cn: the three-tier structure (individual free/Plus, institutional SaaS, on-premises) is confirmed but **no price is published anywhere** — institutional sales are sales-led, and international availability is signaled (English PR push) but not committed. Updated `[[wenge-sciencepro-research-platform]]` with the data-scale moat (270M papers/patents, 32M project records), the 13-agent marketplace, and the state-coordinated distribution contrast with the EHR-embed pattern.
2. **Alexandria live-status** — the Firecrawl self-service payout watch has **not fired**: provider sign-up is still a waitlist. Updated `[[compensated-knowledge-supply-agents]]` with the per-tool pricing mechanics (free discovery, priced retrieval, org-admin terms acceptance) and the Wikimedia Enterprise template — charging for structure and maintenance on freely-licensed content.
**Watches checked, no fires:** Scholar Labs GA (nothing), host native research workspace (nothing), Perplexity sonar-deep-research retirement (scheduled for tomorrow, 2026-09-27).
other finish_cycle
Agent
Cycle complete. Two findings were updated in place (reusing their stable slugs), one open thread was resolved, and all watch tripwires were checked — none fired. The market map now has a firmer answer on SciencePro's commercial model (unpublished, sales-led beyond the free tier) and a current snapshot of the agent-data supply layer's status.