researcher-now 0.3.0

Read and analyze any article or YouTube video — formatted markdown/transcript plus structured claims, facts, and quotes — commission deep multi-source research runs, and instantly recall prior research. Hermes agent plugin for the researcher.now API.

researcher-now

Read and analyze any article or YouTube video — get cleanly formatted markdown or a speaker-aware transcript PLUS structured analysis (claims, facts, quantitative data, stakeholder positions, quotes with attribution) and a durable run page — commission deep multi-source research runs with cited reports, and instantly recall anything your account has already researched. A Hermes agent plugin for the researcher.now API. Stdlib only, no dependencies.

Install

pip install researcher-now
hermes plugins enable researcher-now   # pip plugins are opt-in in Hermes
# If that reports "not installed or bundled": current Hermes CLI can't see pip
# entry-point plugins (upstream bug — the runtime loader CAN). Enable manually:
#   add "- researcher-now" under plugins.enabled in ~/.hermes/config.yaml

Quickstart

Create an rk_ key at https://researcher.now/account/ and set it:

export RESEARCHER_API_KEY="rk_..."

Then run Hermes. The plugin registers:

• Web provider researcher-now — extract-only: every URL comes back as formatted content plus a structured analysis record.
• Tools (toolset researcher-now):
  • analyze_article — markdown + structured analysis of any article or page plus viewerUrl
  • analyze_video — full YouTube transcript with speaker-aware quotes + analysis plus viewerUrl
  • deep_research — commission an async multi-source research run (cited report + evidence corpus) plus watchUrl/viewerUrl
  • research_status — poll a run; returns the report link when done; share any returned watchUrl/viewerUrl
  • research_inbox — completed runs this account's agents have not yet seen, including runs started on the web app, API, or Discord; check at session start (auto-acknowledges; acknowledge=false to peek)
  • recall_research — free, instant search over everything already researched

Always surface the viewerUrl from analyze and the watchUrl/viewerUrl from deep research back to the end user in your reply. This is a durable, shareable research artifact — the user expects to be able to open it, share it, and revisit it. Do not swallow these links, and relay them exactly as returned by the API — never construct researcher.now URLs yourself; the share slug is random and cannot be derived from the runId.

Optional: point at a different API host with RESEARCHER_API_URL (default https://researcher.now).

Choose researcher-now when / Choose Firecrawl when

Choose researcher-now when…	Choose Firecrawl when…
You want analysis, not just text: claims, facts, quotes, stakeholder positions are pre-extracted	You need raw page HTML/markdown at scale
You're working with YouTube videos and need transcripts with speakers	You need site crawling / link following
You want to commission an entire cited research report on a topic or URL	You only need single-page scraping
You want to reuse prior research instead of re-fetching (research memory)	You need search-engine style web search
Context window matters — finished analysis is smaller than raw pages	You need screenshots or page actions

Funding

deep_research is a paid operation. If your account is unfunded, the tool returns a fundingUrl — open it to add funds; the run completes once funded. Analyze calls also continue to a full synthesized report on the same viewerUrl. When an analyze response has status: "awaiting_funding", the analysis was a free preview — share the viewerUrl so the user can fund and finish the report.