Skip to the rows

agentsclimarketplace

Perplexity search

Skill semantic-craft/opencli-stack/skills/perplexity-search

OpenCLI 技能资产与运行时插件适配器的 monorepo（学术检索 + 报销流程）。Monorepo for OpenCLI skill assets and runtime plugin adapters.

Install

npx -y skills add semantic-craft/opencli-stack --skill perplexity-search

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

13 days oldThe repository was created 13 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Web search and quick grounded answers via Perplexity CLI. Use when the user wants to search the web, find current information, get a grounded answer, verify a fact, or says "用 Perplexity 搜", "pplx search", "search for", "look up", "what's the latest on". Returns structured results with URLs or synthesized answers with citations. For multi-source deep research, route to tavily-research when that skill is available.

SKILL.md

3.4 KB, as published. Nobody here has run it

perplexity search

Web search and quick grounded answers via pplx.

When to use

Find current information, recent news, web sources
Get a quick grounded answer with citations
Verify a fact or claim against web sources
First step before deep research: search → research

Not for: reading a specific URL (use Jina/Firecrawl), academic paper discovery (use Jina arXiv/SSRN), deep multi-source investigation (use tavily-research).

Quick start

# Structured search results with URLs
pplx search "EU AI Act implementation timeline" --max-results 5 --json

# Quick grounded answer
pplx ask "What are the current rules on AI liability in the EU?"

# Agent API — let the model refine its own queries
pplx agent "Latest developments in data portability regulation" --preset fast-search

search — structured results

# Recent news
pplx search "AI governance" --recency month --max-results 5 --json

# Domain-restricted
pplx search "AI Act delegated acts" --domain europa.eu --domain eur-lex.europa.eu --json

# Exclude noise
pplx search "data portability" --domain=-reddit.com --domain=-quora.com --json

# Chinese sources
pplx search "人工智能立法" --language zh --max-results 10 --json

# URLs only (pipe to Jina for full text)
pplx search "AI governance frameworks" --max-results 10 --cite-only

# BibTeX for Zotero import
pplx search "platform liability EU" --max-results 20 --bibtex -o seed.bib

ask — synthesized answer

# Basic
pplx ask "Does GDPR Article 20 cover inferred data?" --search-context-size low

# Academic sources
pplx ask "Main scholarly positions on platform liability in EU law" --search-mode academic

# Save JSON for downstream parsing
pplx ask "Is the EU AI Act risk classification adequate for foundation models?" \
  --search-context-size high --json -o claim.json

Defaults

Command	Start with	Escalate to
`search`	`--max-results 5 --json`	10-15 if too few hits
`ask`	`--search-context-size low`	`medium` if answer is thin
`agent`	`--preset fast-search`	`pro-search` if more depth needed

Cross-tool pipelines

# Discover (Perplexity) → Read full text (Jina)
pplx search "topic" --max-results 10 --cite-only \
  | xargs -I {} -P 5 jina read "{}" > sources.md

# Expand query → batch search all reformulations
pplx expand "platform liability EU" -n 6 \
  | pplx batch -P 4 search --max-results 5 --cite-only

Result handling

Cite from --json structure, not from model prose — URLs in text may be paraphrased.
If results are thin or off-target, say so. Don't substitute adjacent results.
No silent supplement — don't fill gaps from training data without flagging.

See also

tavily-research — deep multi-source research
perplexity-cli — full CLI reference

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.