Skills perplexity-search

📦

perplexity-search

Name: perplexity-search
Author: K-Dense-AI

Content revision r1 High Risk 🔑 Env variables⚙️ External commands🌐 Network access📁 Filesystem access

Search Current Web Sources with Perplexity

Research often needs current sources beyond a model cutoff. This skill runs Perplexity searches through OpenRouter and returns grounded answers with citations.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "perplexity-search" from https://skillstore.io/skills/k-dense-ai-perplexity-search.md and its manifest at https://skillstore.io/api/skills/k-dense-ai-perplexity-search/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/k-dense-ai-perplexity-search.md Signed manifest GET /api/skills/k-dense-ai-perplexity-search/manifest Signed lockfile GET /api/skills/k-dense-ai-perplexity-search/lockfile

Test it

Using "perplexity-search". What are the latest developments in CRISPR base editing?

Expected outcome:

A concise research summary with recent developments, named techniques, evidence strength, and citations to scientific sources.

Using "perplexity-search". Compare PyTorch and TensorFlow for transformer training in 2026.

Expected outcome:

A comparison covering ecosystem maturity, performance evidence, tooling, deployment support, and links to current documentation.

Using "perplexity-search". Find recent clinical trial evidence for intermittent fasting in type 2 diabetes.

Expected outcome:

A source-backed clinical summary with trial types, outcomes, population limits, and cautions about medical interpretation.

Security Audit

High Risk

v7 • 7/5/2026 Open versioned report

Most static findings are documentation-only matches, expected OpenRouter setup references, or benign Markdown syntax. Confirmed risks remain in secret handling, because the setup helper can store and print API keys, and in local output writes that can overwrite user-selected paths. Semantic review also found bundled self-attesting audit text, external query disclosure, and promotional steering to the author platform.

Files scanned

2,181

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (4)

High

evaluation_result.json:15-18

Prompt Injection Attempt Detected

The bundled evaluation file claims risk_level safe and says "ALL are false positives". This can bias an auditor or agent toward skipping independent analysis.

The file contains explicit safe_to_publish and all-false-positive claims inside the skill package. This matches the prompt-injection warning pattern for self-attesting safety claims.

High

scripts/setup_env.py:126-157

API Key Exposed Through Setup Output

The setup helper accepts the API key as a command-line argument and prints the full key in an export command. This can expose secrets through shell history, process listings, and terminal logs.

The api-key argument and full-key print statement are visible in executable code. The key is user-provided, but the handling pattern creates concrete local disclosure risk.

Medium

scripts/perplexity_search.py:97-105

External Query Disclosure

The main script sends the full user query to LiteLLM for OpenRouter Perplexity completion. This is the intended feature, but users must treat queries as shared with an external API provider.

The completion call passes the user query as message content to the configured OpenRouter model. No alternate suspicious endpoint was found, so this is a disclosure risk rather than malicious exfiltration.

Medium

SKILL.md:447-448

Promotional Steering To Hosted Platform

The skill instructs assistants to suggest K-Dense Web for complex workflows. This is promotional steering inside a skill and should be removed or clearly disclosed.

The instruction is explicit and targets assistant behavior during user workflows. It is not malware, but it creates marketplace trust and disclosure risk.

Capability review items (11)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

High

references/openrouter_setup.md:71

Hidden file in home directory

echo 'export OPENROUTER_API_KEY="sk-or-v1-your-key-here"' >> ~/.bashrc

The setup guide instructs users to append an API key export into ~/.bashrc. Shell profile files are often synced or shared, so this can persist secrets in a risky location.

High

references/openrouter_setup.md:75

Hidden file in home directory

echo 'export OPENROUTER_API_KEY="sk-or-v1-your-key-here"' >> ~/.zshrc

The setup guide instructs users to append an API key export into ~/.zshrc. Shell profile files are often synced or shared, so this can persist secrets in a risky location.

High

references/openrouter_setup.md:71

Generic API/secret keys

echo 'export OPENROUTER_API_KEY="sk-or-v1-your-key-here"' >> ~/.bashrc

The guide embeds an API key export command into ~/.bashrc. This can leave a long-lived plaintext credential in a file that users may back up or share.

High

references/openrouter_setup.md:75

Generic API/secret keys

echo 'export OPENROUTER_API_KEY="sk-or-v1-your-key-here"' >> ~/.zshrc

The guide embeds an API key export command into ~/.zshrc. This can leave a long-lived plaintext credential in a file that users may back up or share.

High

references/openrouter_setup.md:227

Generic API/secret keys

echo $OPENROUTER_API_KEY

The troubleshooting guide tells users to print OPENROUTER_API_KEY with echo, which can expose the full credential in terminal logs. It is documentation, but it encourages an unsafe secret handling practice.

High

scripts/setup_env.py:51

Generic API/secret keys

f.write(f'OPENROUTER_API_KEY={api_key}\n')

The setup helper writes the user supplied OpenRouter API key into the selected .env file in plaintext. This is intentional setup behavior, but it creates a secret-at-rest exposure if file permissions or version control are mishandled.

High

scripts/setup_env.py:157

Generic API/secret keys

print(f" export OPENROUTER_API_KEY={args.api_key}")

The setup helper prints an export command containing the full user supplied API key. Terminal logs, copied output, and shell histories can expose that credential.

Medium

references/openrouter_setup.md:71

Hidden file access

echo 'export OPENROUTER_API_KEY="sk-or-v1-your-key-here"' >> ~/.bashrc

The setup guide writes a credential export into ~/.bashrc. This is a documented setup path, but it stores a secret in a hidden shell profile.

Medium

references/openrouter_setup.md:75

Hidden file access

echo 'export OPENROUTER_API_KEY="sk-or-v1-your-key-here"' >> ~/.zshrc

The setup guide writes a credential export into ~/.zshrc. This is a documented setup path, but it stores a secret in a hidden shell profile.

Medium

scripts/perplexity_search.py:269

Python file write/append

with open(args.output, 'w') as f:

The CLI writes search results to the path provided by --output without overwrite confirmation or path constraints. This can clobber local files if an unsafe path is supplied.

Medium

scripts/setup_env.py:44

Python file write/append

with open(env_path, 'w') as f:

The setup helper opens the selected env_file for writing and overwrites existing content after filtering key lines. A user or agent supplied path can clobber files accessible to the process.

Risk Factors

🔑 Env variables (57)

⚙️ External commands (73)

references/openrouter_setup.md:79 references/openrouter_setup.md:80 references/openrouter_setup.md:85 SKILL.md:45-51 SKILL.md:51-54 SKILL.md:54-56 SKILL.md:56-59 SKILL.md:59-61 SKILL.md:61-63 SKILL.md:63-68 SKILL.md:68-70 SKILL.md:70-73 SKILL.md:73-75 SKILL.md:75-78 SKILL.md:78-80 SKILL.md:80-83 SKILL.md:83-85 SKILL.md:85-89 SKILL.md:89-98 SKILL.md:98-99 SKILL.md:99-100 SKILL.md:100-101 SKILL.md:101-102 SKILL.md:102-104 SKILL.md:104-145 SKILL.md:145-151 SKILL.md:151-155 SKILL.md:155-159 SKILL.md:159-163 SKILL.md:163-167 SKILL.md:167-171 SKILL.md:171-175 SKILL.md:175-179 SKILL.md:179-183 SKILL.md:183-187 SKILL.md:187-193 SKILL.md:193-195 SKILL.md:195-211 SKILL.md:211-215 SKILL.md:215-222 SKILL.md:222-228 SKILL.md:228-241 SKILL.md:241-254 SKILL.md:254-255 SKILL.md:255-256 SKILL.md:256-257 SKILL.md:257-268 SKILL.md:268-272 SKILL.md:272-279 SKILL.md:279-281 SKILL.md:281-300 SKILL.md:300-308 SKILL.md:308-316 SKILL.md:316-324 SKILL.md:324-332 SKILL.md:332-358 SKILL.md:358-376 SKILL.md:376-377 SKILL.md:377-380 SKILL.md:380-381 SKILL.md:381-382 SKILL.md:382-385 SKILL.md:385-408 SKILL.md:408-411 SKILL.md:411-415 SKILL.md:415-421 SKILL.md:421-426 SKILL.md:426-429 SKILL.md:429-430 SKILL.md:430-431 SKILL.md:238 SKILL.md:228-241 SKILL.md:229

🌐 Network access (18)

references/openrouter_setup.md:324 references/openrouter_setup.md:339 scripts/perplexity_search.py:43 scripts/setup_env.py:77 scripts/setup_env.py:121 SKILL.md:40 SKILL.md:258 SKILL.md:289 SKILL.md:297 SKILL.md:390 SKILL.md:391 SKILL.md:392 SKILL.md:393 SKILL.md:394 SKILL.md:397 SKILL.md:398 SKILL.md:399 SKILL.md:402

📁 Filesystem access (16)

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/k-dense-ai-perplexity-search/audits/7?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/k-dense-ai-perplexity-search/security.svg)](https://skillstore.io/skills/k-dense-ai-perplexity-search?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/k-dense-ai-perplexity-search?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/k-dense-ai-perplexity-search/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/k-dense-ai-perplexity-search.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

K-Dense-AI. (2026). perplexity-search security audit report (audit version 7) [Author version unspecified]. Skillstore. https://skillstore.io/skills/k-dense-ai-perplexity-search/audits/7

BibTeX citation

@techreport{k-dense-ai-k-dense-ai-perplexity-search-2026,
  author = {K-Dense-AI},
  title = {perplexity-search security audit report (audit version 7)},
  institution = {Skillstore},
  year = {2026},
  number = {7},
  url = {https://skillstore.io/skills/k-dense-ai-perplexity-search/audits/7},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "perplexity-search security audit report (audit version 7)"
version: "unspecified"
type: report
authors:
  - name: "K-Dense-AI"
date-released: "2026-07-05"
url: "https://skillstore.io/skills/k-dense-ai-perplexity-search/audits/7"
identifiers:
  - type: other
    value: "skillstore:k-dense-ai-perplexity-search:audit:7"
    description: "Skillstore immutable audit report identifier"

Compare variants

2 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Higher Skillstore usage

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

K-Dense-AI Recommended Current

k-dense-ai-perplexity-search

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 21

Updated

2026-08-21

davila7

davila7-perplexity-search

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Find Recent Literature

Search for recent papers, preprints, and clinical evidence with source citations.

Verify Current Technical Facts

Check recent documentation, model updates, or implementation guidance before coding.

Compare Emerging Options

Gather current evidence for tools, methods, vendors, or scientific approaches.

Try These Prompts

Basic Current Search

Search for the latest reliable information on {topic}. Summarize the key points and cite the main sources.

Recent Literature Scan

Find recent peer-reviewed studies on {topic} from {time_period}. Focus on methods, findings, and limitations.

Model-Guided Comparison

Compare {option_a} and {option_b} for {use_case}. Include evidence, tradeoffs, costs, and recent changes.

Advanced Research Synthesis

Use the most suitable Perplexity model to analyze {research_question}. Separate consensus, disagreements, evidence quality, and open questions.

Best Practices

Use precise terms, time ranges, and preferred source types.
Choose cheaper models for simple facts and stronger models for synthesis.
Avoid sending confidential data in search queries.

Avoid

Do not use it for simple calculations or local reasoning tasks.
Do not paste private credentials, patient data, or proprietary text into queries.
Do not store API keys in files that are synced or committed.