Browser Use
All site skills

Research and public data

ods.od.nih.gov

NIH ODS - public fact-sheet capture

Notes an AI agent reads before it works on ods.od.nih.gov: what to open, what to call, what breaks. Written by browser-harness contributors after real runs, and kept on GitHub under the MIT license.

Run it in Browser Use CloudSource on GitHub

Give it to your agent

Any coding agent

Start the task with this line. The Markdown version is the same notes without the page around them.

Read https://browser-use.com/skills/ods.md, then <your task>

Browser Harness

Save the files into your workspace and turn domain skills on. The code below calls its helpers, such as http_get and js. What Browser Harness is.

d="${BH_AGENT_WORKSPACE:-$HOME/.config/browser-harness/agent-workspace}/domain-skills/ods"
mkdir -p "$d" && cd "$d"
curl -sLO https://raw.githubusercontent.com/browser-use/browser-harness/main/agent-workspace/domain-skills/ods/fact-sheets.md
export BH_DOMAIN_SKILLS=1

Observed on ods.od.nih.gov in September 2026.

  • Fact-sheet URLs use /factsheets/<Nutrient>-HealthProfessional/ and /factsheets/<Nutrient>-Consumer/. Confirm the audience in the visible title.
  • Try ordinary HTTP first. A plain request returned 403 while the same public page rendered normally in Chrome. Do not archive a denial body as the fact sheet; a normal browser visit may provide the public article without login.
  • The rendered article contains the sheet, references, disclaimer and displayed update date. #fact-sheet contains the substantive sections. If capturing DOM, label it rendered article HTML, not the original HTTP response.
  • Locate sections by visible heading text, then use that heading's observed ID. Numeric IDs such as h18 are specific to a sheet's organization, not universal identifiers for the same topic on every nutrient page.
  • Some headings have duplicate generated IDs (heading-0, etc.). Prefer the actual heading element within the article rather than a global ID match.
  • Inline references include tooltip citation text in the DOM. Naive textContent or static HTML-to-text extraction can repeat full citations inside paragraphs. Keep the original capture, and distinguish displayed prose from tooltip text before doing content-length checks or extracting sentences.
  • Record the article's Updated date separately from retrieval time. A fresh capture does not mean the scientific content was updated that day.

The site's linked source records and original publications still need their own identity and scope checks; the fact-sheet capture is not review of every cited paper. No account access or form submission is needed for this workflow.

fact-sheets.md · updated Sep 6, 2026 · 256 words

Sites change. These notes describe what worked when they were last updated, not a guarantee they still do. If something broke, the fix is a pull request to browser-harness.

More research and public data skills