Borrowing it
Nothing to install: this file belongs to vvzvlad/research-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/vvzvlad/research-mcp/main/AGENTS.mdgit clone --depth 1 https://github.com/vvzvlad/research-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/instructions/vvzvlad/research-mcp/agents-md)<a href="https://agentmods.dev/instructions/vvzvlad/research-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/vvzvlad/research-mcp/agents-md/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/instructions/vvzvlad/research-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/vvzvlad/research-mcp/agents-md.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 | $0.01633 | $0.01633 |
| Opus 5.5 | $0.00653 | $0.00653 |
| Sonnet 5.5 | $0.00327 | $0.00327 |
| Haiku 4.5 | $0.00163 | $0.00163 |
Grade A, and why
research-mcp AGENTS.md 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 21d 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 — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Instructions — research-mcp
An MCP "facade" that hides a pyramid of search/read providers behind a single
streamable-http MCP endpoint (/mcp) and exposes 4 tools: web_search,
read_page, read_pages, search_and_read. No auth in the app: Traefik +
basicAuth on the host handles it.
Plugin architecture (types + instances)
- A type is an implementation class (
src/providers/<type>.py), registered with@register("type")→REGISTRY(src/providers/registry.py). - An instance is a configured copy of a type whose secrets/URL come from named ENV variables. Several instances of one type are allowed.
- Which instances exist and the pipeline order live in code
(
src/pipeline_config.py:INSTANCES,SEARCH_PIPELINE,READ_PIPELINE); the loader insrc/pipeline.pyresolves the names withos.getenvand enables an instance only when its required variables are set. duckduckgo(search) andtrafilatura(read) need no ENV var: the server must build and both tools must work on an empty environment.duckduckgosteps aside whenSEARXNG_URLis set (Instance.disabled_by_env): our SearXNG queries the same DuckDuckGo from the same IP.- A
UrlSpecificReader(youtube,supadata,instagram,instagram_profile,scholarly) is offered only the urls itaccepts, before the PDF probe; its answer is final, a failure falls through to the rest of the read chain unless the reader setsfalls_through = False(the YouTube readers): then the read fails once the url-specific readers are spent. A reader withfetches_target_locally = True(scholarly) gets the SSRF-guarded client, liketrafilatura.
Project structure
src/server.py—build_server()with the 4@mcp.tooldefinitions.src/pipeline.py— instance loader,ClientManager(one httpx client per proxy URL), search (concurrent, merge/dedup/rerank), read (PDF probe + fallback chain) andsearch_and_read.src/pipeline_config.py— instances and pipeline order.src/providers/—base.py(interfaces,ProviderError,ProviderConfig),_http.py(shared retry and 402/429 → failover policy),_url_guard.py(SSRF guard),pdf.py, and one module per type.src/rerank.py— optional Jina rerank of merged search results.src/quota.py— persistent per-instance quotas and billed counters (SQLite).src/failure_reason.py—classify(exc)→ failure category.src/formatting.py— pure renderers of the LLM-facing results and status lines.src/settings.py— non-secret knobs only, all defaulted.main.py— entry point: stderr + file log sink, build server, run.data/— runtime state (research-mcp.log,quota.sqlite3); gitignored, a docker volume in production.tests/— pytest, network mocked with respx.server.json— MCP registry metadata.
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.
- 21d ago Changed · +8 lines · +147 tokens per session 7112399770ae
- 23d ago Changed · +2 lines · +39 tokens per session 988efa9405d9
- 24d ago Changed · +2 lines · +36 tokens per session bb596412887d
- 26d ago Changed · -46 lines · -1,020 tokens per session bf587b2cfd74
- 1mo ago Changed · +9 lines · +147 tokens per session 80dd3882c2b9
- 1mo ago Changed · +14 lines · +283 tokens per session e28b33609a6c
- 1mo ago Changed · +5 lines · +102 tokens per session 7ff036e95144
- 1mo ago Changed · +30 lines · +607 tokens per session b1093b30a643
- 2mo ago First seen · 86 lines · 1,292 tokens per session scan A e1a892453f8c
research-mcp AGENTS.md is an instructions file published in the GitHub repository vvzvlad/research-mcp (0 stars, last pushed 22d ago), licensed MIT. It adds 1,633 tokens to every session, about $0.0065 per session on Opus 5.5. 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 2026-08-31.
research-mcp AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| research-mcp AGENTS.mdvvzvlad/research-mcp | Instructions file | 0 | 22d ago | MIT | A | 1,633 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 274k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |