ReddRadar
Agent skill
browse-now
Control the user's actual browser through the browse-now CLI when a task needs authenticated pages, dynamic interaction, form filling, screenshots, or other browser automation that web search cannot reach. Prefer it whenever the task depends on the user's real browser state rather than a generic fetch.
Install this agent skill to your Project
npx add-skill https://github.com/nowledge-co/community/tree/main/nowledge-mem-browse-now-npx-skills/skills/browse-now
SKILL.md
Browse Now
Use the user's real browser when the task needs live interaction, login state, or dynamic content.
When To Use
Use browse-now when:
- the site requires the user's existing browser session or login cookies
- the task needs clicking, typing, scrolling, or navigating through a live page
- the page is dynamic enough that plain HTTP fetches or web search will miss the real UI state
- the user needs a screenshot or content extraction from the rendered page
Skip it when ordinary web search or direct HTTP fetching would answer the question faster.
browse-now is local-only. It must run on the same machine as the Nowledge Mem app, the browser, and the Exchange extension. It is not exposed through Access Anywhere.
Core Loop
browse-now open <url>
browse-now snapshot -i
browse-now click @e5
browse-now wait 2
browse-now get url
browse-now snapshot -i
Preferred Interaction Order
- Start with
browse-now snapshot -ito get interactive refs. - Use
click @eNandfill @eN ...as the primary interaction path. - After navigation or major DOM changes, run
snapshot -iagain because refs reset. - Confirm page changes with
browse-now get urlorbrowse-now get title. - If accessibility data is sparse, use
browse-now find "query",click -T "text", orscreenshot.
Practical Commands
browse-now open https://example.com
browse-now snapshot -i
browse-now find "login button"
browse-now click @e12
browse-now fill @e3 "search text" --submit
browse-now get page-text --max-chars 4000
browse-now screenshot /tmp/page.png
Reliability Notes
- Refs from
snapshot -iare the most reliable click targets. - Text clicks are a fallback for dialogs, dropdowns, or poor accessibility trees.
- A successful click does not bypass login walls, paywalls, or anti-bot behavior.
- Always verify where you landed before assuming the browser action failed.
Requirements
- Nowledge Mem desktop app installed
browse-nowavailable onPATH- Nowledge Mem Exchange extension installed in a supported browser
- A connected browser session the agent can control
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
nowledge-co/community
search-memory
Search your personal knowledge base when past insights would improve response. Recognize when stored breakthroughs, decisions, or solutions are relevant. Search proactively based on context, not just explicit requests.
nowledge-co/community
check-integration
Check Nowledge Mem setup, detect your agent, and guide native plugin installation. Use when the user asks about setup, configuration, or when memory tools aren't working as expected.
nowledge-co/community
distill-memory
Capture breakthrough moments and valuable insights as searchable memories in your knowledge base.
nowledge-co/community
read-working-memory
Read your daily Working Memory briefing to understand current context. Contains active focus areas, priorities, unresolved flags, and recent knowledge changes. Load this automatically at the beginning of sessions for cross-tool continuity.
nowledge-co/community
status
Check Nowledge Mem connection status, server version, CLI version, and configuration. Use when diagnosing issues or verifying setup.
nowledge-co/community
save-handoff
Save a concise handoff summary only when the user explicitly requests it. Use this for resumable progress notes in generic agent environments where a real session importer is not guaranteed.
Didn't find tool you were looking for?