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 sydasif/web-search-mcp --skill researchgit clone --depth 1 https://github.com/sydasif/web-search-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/sydasif/web-search-mcp/research)<a href="https://agentmods.dev/skills/sydasif/web-search-mcp/research"><img src="https://agentmods.dev/badge/skills/sydasif/web-search-mcp/research/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/sydasif/web-search-mcp/research"><img src="https://agentmods.dev/badge/skills/sydasif/web-search-mcp/research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00026 | $0.03511 |
| Opus 5 | $0.00013 | $0.01755 |
| Sonnet 5 | $0.00005 | $0.00702 |
| Haiku 4.5 | $0.00003 | $0.00351 |
Grade A, and why
research 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 6d 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.
How it starts
The opening of the file, as written. The whole thing — 336 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deep Research
Operating Principle
Research is breadth-first, then depth-first. Start wide to map the landscape, then dive into the highest-signal sources. Do not jump to a single source type and call it done — each platform reveals a different facet of the topic.
Good research answers not just what happened, but who is saying it, how confident the evidence is, and what the competing narratives are.
Cross-source corroboration: A claim found in 3+ independent sources is stronger than any single source. Multi-platform coverage is the highest-confidence signal.
Internalize the research first: Ground your synthesis in the ACTUAL research content, not your pre-existing knowledge. If sources talk about "ClawdBot" and you assumed "Claude Code", do not conflate them.
Important: What the MCP Handles vs What You Do
This skill uses the web-search MCP server. Search, dedup, clustering, and stats computation, here you do all that synthesis work manually as you read tool results.
The MCP gives you:
- Search and retrieval across 10 tools covering web, social, academic, and developer platforms
- Clean extracted content from URLs
- Rich engagement data (upvotes, comments, views) from community platforms
You are responsible for:
- Running searches sequentially (no engine to fan out in parallel)
- Counting and tracking which claims came from which source
- Spotting patterns and clusters as you read
- Computing stats manually from result sets
- Noticing "X Reply Clusters" in Reddit threads yourself
Cost of comparison mode: Comparing 2 entities means roughly 2x the tool calls (~8-12 per entity). Plan your turn budget accordingly. For 3-way comparisons, consider doing a single focused comparison instead of exhaustive per-entity research.
Tool Reference — 10 Web Search Tools
Tier 1 — Broad Discovery & Reference
| Tool | What It Does | When To Use |
|---|---|---|
mcp__plugin_web-search_web-search_search_web |
DuckDuckGo or Exa web/news search (via provider param) |
First pass: news, background, official sources |
mcp__plugin_web-search_web-search_search_web (domain) |
DuckDuckGo (site:) or Exa (include_domains) scoped to a domain | Targeted docs: docs.python.org, react.dev, RFCs |
mcp__plugin_web-search_web-search_fetch_page |
Clean HTML-to-markdown extraction from URLs | Read articles, changelogs, specs, papers |
mcp__plugin_web-search_web-search_search_wikipedia |
Full article text via MediaWiki API | Factual summaries, background research, citations |
mcp__plugin_web-search_web-search_search_arxiv |
Academic paper search w/ Lucene field prefixes | Research papers, literature reviews, citations |
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.
- 6d ago Changed · +1 lines 59f89bb53160
- 10d ago First seen · 335 lines · 26 tokens per session scan A 634af1c4a6d6
research is a skill published in the GitHub repository sydasif/web-search-mcp (24 stars, last pushed 5d ago), licensed MIT. It adds 26 tokens to every session and 3,511 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
opencli-usage
Use when running OpenCLI commands to interact with websites (Bilibili, Twitter, Reddit, Xiaohongshu, etc.), desktop apps (Cursor, Notion), or public APIs (HackerNews, arXiv). Covers installation, command reference, and output formats for 100+ adapters.
qiaomu-opencli-usage
Use when running OpenCLI commands to interact with websites (Bilibili, Twitter, Reddit, Xiaohongshu, etc.), desktop apps (Cursor, Notion), or public APIs (HackerNews, arXiv). Covers installation, command reference, and output formats for 79+ adapters.
mk:wiki-research
Fetch external sources (web / arXiv / GitHub) and turn them into scanner-gated wiki CANDIDATES — never canonical pages. Every fetched byte is DATA; url-guarded, size-capped, redirect-re-validated, injection-scanned, and secret-scrubbed before a candidate is even created. Requires network. NOT for capturing local…
social-reddit-card
Realistic Reddit post card with vote rail and comment count, suited to video overlays or story sharing.
issue
Use when starting a chain from a GitHub issue — turning an issue URL or number into a triaged, planned, dispatched, and reviewed pull request. Classifies the thread (bug → root-cause discipline, feature → plan chain, question → drafted reply), synthesizes a spec from the issue's own acceptance criteria, then runs the…
ultralytics-platform
This skill should be used when user asks to "upload my model to Ultralytics Platform", "push this run to the platform", "upload a dataset to platform", "download a dataset from platform", "search platform datasets", "start cloud training", "train on platform GPUs", "export a model on platform", "deploy a model…