01
It starts by itself
every weekday morning. Runs in Claude Code routine or cron.
- ↻ every weekday morning
0 8 * * 1-5
parallel-web · 14K installsParallel Web Search skill
Runs by itself, every weekday morning.
Cited web answers every morning.
Install the Parallel Web Search skill and make it a habit: follow https://gethabit.dev/skills/parallel-web/parallel-web-search.md
- 1Copy
- 2Paste into Claude Code, Codex or Cursor
- 3It runs by itself
14K installs90 starsMIT licenceSource on GitHub ↗
Parallel Web Search
by parallel-web · 14K installs
The skill
The parallel web search skill runs parallel-cli search from the command line, saves the full results as JSON and writes an answer with an inline citation for every claim.
Only when you ask.
The habit
Cited web answers every morning.
↻ Runs by itselfevery weekday morning
How the habit works.
You don't have to.
When does it start? How does it check its work? What does it remember?
One run, step by step.
Then it stops.
One question per run, with at most one refined retry when results are sparse.
- 01
Read the log
The agent opens the run log to see which questions ran before and where their results live.
- 02
Name the file
It picks a short lowercase file name with hyphens that fits today's question.
- 03
Search
It calls parallel-cli search with the objective, a few -q keywords, --json, --max-results 10 and -o.
- 04
Check the result
It reads the saved JSON, then checks the exit status, API errors and warnings.
- 05
Write the answer
It leads with the finding and links each fact to a URL from the JSON.
- 06
List sources
It adds a Sources section, names the JSON file path and appends a line to the run log.
What it may do.
Nothing more.
It writes only in its own folder and asks you before anything goes out.
- Writes only in/tmp/*.json, loops/parallel-web-search-loop/**
- Talks tono outside service
- Sendsnothing
- Your OKNothing goes out, so it needs no OK.
Is it for you?
Use it when it fits.
Read both lists. If the left one sounds like you, the Parallel Web Search skill helps.
Use it when
- ✓ you need sourced answers with links
- ✓ you want results saved as JSON
- ✓ no web search MCP tool is available
Skip it when
- – the Parallel web_search MCP tool already works
- – you need deep, exhaustive research
Parallel Web Search skill, answered.
What happens when the search fails?
The run reports the error and stops. An old JSON file does not count as a success.
Do I need anything installed?
Yes. The loop needs parallel-cli with a working login. The skill points to its setup command when the CLI is missing.
Can I limit the sources?
Yes. Add --include-domains or --exclude-domains to the saved question, or --after-date for recent news.
How do I install the Parallel Web Search skill?
Copy the prompt at the top into your agent. It installs the skill from parallel-web/parallel-agent-skills and sets up the habit. Or run npx skills add parallel-web/parallel-agent-skills --skill parallel-web-search.
Is the Parallel Web Search skill free?
Yes. It is published under MIT. Habit is free too.
Parallel Web Search. Now as a habit.
Install the Parallel Web Search skill and make it a habit: follow https://gethabit.dev/skills/parallel-web/parallel-web-search.md
All parallel-web skills →More notes skills →
Works in
- Claude Code
- Codex
- Cursor
- ChatGPT
- Claude app
Made by parallel-web, published under MIT. Habit links to the original and never changes it. Is this your skill?