Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add devdotbo/perplexity-web-mcp --skill datagit clone --depth 1 https://github.com/devdotbo/perplexity-web-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/devdotbo/perplexity-web-mcp/data)<a href="https://agentmods.dev/skills/devdotbo/perplexity-web-mcp/data"><img src="https://agentmods.dev/badge/skills/devdotbo/perplexity-web-mcp/data/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/devdotbo/perplexity-web-mcp/data"><img src="https://agentmods.dev/badge/skills/devdotbo/perplexity-web-mcp/data.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00122 | $0.03628 |
| Opus 5.5 | $0.00049 | $0.01451 |
| Sonnet 5.5 | $0.00024 | $0.00726 |
| Haiku 4.5 | $0.00012 | $0.00363 |
Grade A, and why
perplexity-web-mcp scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
100% identical to perplexity-web-mcp — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 350 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Perplexity Web MCP
Search the web and query premium AI models through Perplexity AI.
Quick Reference
Run pwm --ai for comprehensive AI-optimized documentation covering all
commands, models, MCP tools, auth flows, and error recovery.
pwm --ai # Full AI reference (RECOMMENDED first step)
pwm --help # CLI help
pwm login --check # Check auth status
Critical Rules
- Authenticate first: Run
pwm loginbefore any queries - Tokens last ~30 days: Re-run
pwm loginon 403 errors - Default to
claude_opuswith--thinking --intent detailed-- always use the best model - Always output to file -- pipe to
/tmp/pwm-*.txt, never dump into conversation context - Never use Deep Research autonomously -- only when the user explicitly asks
Default Invocation
Every pwm ask call uses these flags -- no exceptions:
pwm ask "question" -m claude_opus --thinking --intent detailed > /tmp/pwm-TASKID.txt
-m claude_opus-- always explicit model, never barepwm askor auto/sonar--thinking-- always enabled--intent detailed-- always detailed (the standard research tier)- Output piped to
/tmp/pwm-*.txt-- never dump raw content into conversation context - No
-sflag by default -- no additional source focus unless the query specifically needs it - No
--jsonby default -- raw markdown is cleaner for file output. Add--jsononly when programmatic parsing is needed.
Rate Limiting and Parallelism
0.5 requests per second = 1 request every 2 seconds. When launching parallel requests, stagger by sleep 2 between each launch. Parallel execution is the default for independent queries.
# Parallel pattern with 2s stagger
pwm ask "Q1" -m claude_opus --thinking --intent detailed > /tmp/pwm-q1.txt &
sleep 2
pwm ask "Q2" -m claude_opus --thinking --intent detailed > /tmp/pwm-q2.txt &
sleep 2
pwm ask "Q3" -m claude_opus --thinking --intent detailed > /tmp/pwm-q3.txt &
wait
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 350 lines · 122 tokens per session scan A 2dd8e89a2f4c
perplexity-web-mcp is a skill published in the GitHub repository devdotbo/perplexity-web-mcp (20 stars, last pushed 5mo ago), licensed MIT. It adds 122 tokens to every session and 3,628 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. It is 100% identical to perplexity-web-mcp, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
harbor-daytona
Use Harbor's Daytona sandbox platform for computer use — creating sandboxes, taking screenshots, sending mouse/keyboard input, and building agent loops. Use when the user wants to interact with a GUI, automate a desktop, do computer use, control a browser visually, or run Claude computer use against a Daytona sandbox.
boost-modules
Skill "boost-modules" from av/harbor, covering harbor boost custom modules, module structure, quick reference, output methods and stream text to client.
agentbox-setup
Generate an agentbox.yaml for the current AgentBox workspace. Invoke when the user opens a sandbox without an agentbox.yaml or asks to (re)configure one.
docker
Use when authoring or auditing a Dockerfile, shrinking a bloated image, hardening a container that runs as root, picking a base image, or wiring a Compose dev loop with hot reload. NOT CI builds or deploy-to-host (that is deployment), NOT k8s autoscaling (that is scaling), NOT app-level injection or secrets-in-code…
kastell-ops
Kastell CLI patterns, architecture, anti-patterns, and decision trees. Use automatically when working in Kastell codebase or when asked about Kastell server infrastructure, security audit, hardening, lock, provision, or provider management.
kastell
Server security auditing, hardening, and fleet management. 457 security checks across 30 categories (SSH, Firewall, Docker, TLS, HTTP Headers), CIS/PCI-DSS/HIPAA compliance mapping, 24-step production hardening, and 13 MCP tools. Supports Hetzner, DigitalOcean, Vultr, and Linode with Coolify, Dokploy, and bare VPS…