websearch
Skill0 starsChanged 3 months agoApache-2.0View on GitHub
Search the web with ait-search (search-cli) instead of ad-hoc web search tools - every page found is kept locally, a repeated query marks what is new, every page can carry tags, a status and a note. Use when the user asks to find something on the web, "what do people write about", "find posts", or when a task needs a web sweep.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- PrimeIntellect-ai/prime-agenttheir own version, first seen Oct 8, 2026
- wangbooth/skillstheir own version, first seen Aug 14, 2026
- Elnora-AI/elnora-websearch-toolstheir own version, first seen Jul 14, 2026
- dekoza/oh-my-sloptheir own version, first seen Mar 19, 2026
- roryyu/websearch-free-skilltheir own version, first seen Mar 18, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- PrimeIntellect-ai/prime-agent★ 23ksame name
- wangbooth/skills★ 7same name
- dekoza/oh-my-slop★ 4same name
- roryyu/websearch-free-skill★ 2same name
- Elnora-AI/elnora-websearch-tools★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 24
Look at what the store already holds before asking - often the page is there:
- Distinctive: line 44
Ask several narrow questions rather than one wide one - each is one call, and its results are kept, so the next query only adds.
- Very specific: line 47
Use --site for one surface at a time and --since for the window.
“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Web search with memory
The tool is ait-search; ait-search --help and the repository's SEARCH-API.md
describe the commands. This skill is about how to search well with it.
Before spending: what is configured
ait-search status # engines, keys, prices
ait-search status gemini --probe # does the key work, is the quota exhausted right nowAn error is {"error": code, "message", "hint"}: not_configured means set
the variable, quota_exhausted means put another engine first
(--engine keenable), unsupported on fetch means --format full.
Store first, engine second
Look at what the store already holds before asking - often the page is there:
ait-search history # was this asked already, and when
ait-search grep '"work stealing"' # local full-text over everything seen
ait-search doc list --site github.com --since 14dThen search. A page seen before is the same document, and only never-seen
results carry "new": true. Re-running yesterday's query is the intended way
to track a topic. --wide asks every configured engine and fuses the
rankings; --fetch N brings the top N bodies in the same call. AI engines
(gemini, tavily) return an answer too - read it as one more source with a
score, not as the verdict.
Write the query as a description
Semantic engines (keenable, exa) match meaning: "article by a maintainer comparing Rust async runtimes on tail latency" finds more than "async runtime comparison". Keyword engines (brave, tavily, searxng) want terms. Ask several narrow questions rather than one wide one - each is one call, and its results are kept, so the next query only adds.
Use --site for one surface at a time and --since for the window. An empty
answer says nothing about the surface: an index is a sample, and it lags the
page by days.
Read before judging
A snippet is not the page:
ait-search fetch <id>
ait-search doc get <id> --bodyWrite back what you learned
The store is only worth more than the engine if the reading is recorded:
doc status <id> useful|noise|read|answered- the verdictdoc tag <id> ...- the axes to slice by later (topic, surface, actor)doc note <id> "..."- the one line that saves reopening the page
Report to the user what is new, not the whole list.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.