exa-contents
Skill1 starsChanged 4 months agoMITView on GitHub
Use when an agent already has URLs and needs page text, highlights, summaries, freshness controls, subpages, links, or per-URL statuses through the native stateless exa CLI.
- Reads credentials
2 other projects use a version of this.
Adopted by
Most recently taken up by
- MoizIbnYousaf/marketing-clitheir own version, first seen Nov 1, 2026
- exa-labs/agent-skillstheir own version, first seen Sep 12, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- exa-labs/agent-skills★ 67same name
- MoizIbnYousaf/marketing-cli★ 33same 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.
In this repository2 other files it gives its agents.
- build-with-exaSkill
- exa-searchSkill
The file
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 3 are distinctive, and 9 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
9
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Never place the credential in command arguments, input JSON, source code, or logs.
- Very specific: line 42
Use --input <file|-> for highlights, summaries, freshness, subpages, or extras.
- Very specific: line 58
Do not mix --input with positional URLs or --max-characters.
- Distinctive: line 76
Read Contents input for freshness, subpages, extras, and advanced field shapes.
- Very specific: line 85
Do not treat process exit 0 as proof that every requested URL was extracted.
- Very specific: line 96
Do not wrap text, highlights, or summary in a contents object.
- Very specific: line 97
Do not read results or statuses at the envelope root.
- Very specific: line 98
Do not assume every URL succeeded because the command exited 0.
- Very specific: line 99
Do not mix --input with convenience arguments.
- Very specific: line 100
Do not send stream.
- Distinctive: line 101
Do not request multiple content modes without a concrete need.
- Very specific: line 102
Do not use a huge unbounded text request when a smaller maxCharacters budget suffices.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Exa Contents CLI
Overview
The Contents surface is exa fetch. Use it when URLs are known and the task needs page extraction without another search.
Set the credential in the environment:
export EXA_API_KEY="your_api_key_here"Never place the credential in command arguments, input JSON, source code, or logs.
Quick Start
Fetch one or more URLs:
exa fetch \
"https://example.com/report" \
"https://example.org/analysis"Convenience mode requests text with text.maxCharacters: 3000. Override the per-page text budget:
exa fetch "https://example.com/report" \
--max-characters 5000 \
--timeout-ms 30000The command emits one JSON envelope. Read extracted pages from data.results.
Advanced Input
Use --input <file|-> for highlights, summaries, freshness, subpages, or extras. - reads one complete request object from stdin:
exa fetch --input - <<'JSON'
{
"urls": ["https://example.com/report"],
"highlights": {
"query": "methodology and findings",
"maxCharacters": 1500
},
"maxAgeHours": 24,
"livecrawlTimeout": 12000
}
JSONDo not mix --input with positional URLs or --max-characters. Global flags such as --timeout-ms, --verbose, and --pretty remain valid.
Unknown fields are forwarded. The CLI validates an object; the upstream service remains authoritative for field combinations.
Contents does not support streaming. Omit stream.
Choose One Content Mode
On Contents input, text, highlights, and summary are top-level fields:
| Mode | Use |
highlights | Compact source excerpts for agents and factual lookups |
text | Broad page context, bounded with maxCharacters |
summary | Exa-side per-page compression or structured extraction |
Choose one by default. Multiple modes return overlapping views and may add latency or synthesis work.
Read Contents input for freshness, subpages, extras, and advanced field shapes.
Inspect Every URL Status
A successful envelope can contain a mixture of successful and failed URLs. Always inspect both:
data.resultsfor returned page content.data.statusesfor per-URL success or error states.
Do not treat process exit 0 as proof that every requested URL was extracted. Read Contents output for the response contract and retry behavior.
Search or Fetch?
- Need discovery or ranking: use
exa search. - Already know the URLs: use
exa fetch. - Need compact excerpts while searching: request Search
contents.highlights; do not add a redundant Fetch call. - Need a second extraction policy for selected Search results: pass those URLs to a new Fetch invocation.
Critical Mistakes
- Do not wrap
text,highlights, orsummaryin acontentsobject; that nesting belongs to Search. - Do not read
resultsorstatusesat the envelope root. - Do not assume every URL succeeded because the command exited
0. - Do not mix
--inputwith convenience arguments. - Do not send
stream. - Do not request multiple content modes without a concrete need.
- Do not use a huge unbounded text request when a smaller
maxCharactersbudget suffices.
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.

