Integrations¶

neuroflow connects to four MCP (Model Context Protocol) servers that Claude Code launches automatically via npx. One requires credentials; three work out of the box.

MCP servers¶

Server	npm package	Requires credentials
PubMed / bioRxiv	`paper-search-mcp-nodejs`	❌ None
Miro	`@k-jarzyna/mcp-miro`	✅ `MIRO_ACCESS_TOKEN`
Context7	`@upstash/context7-mcp`	❌ None

All are started automatically by Claude Code — you do not need to run them manually.

Optional — Zotero: neuroflow does not bundle a Zotero server, but if you add a community one (e.g. zotero-mcp — claude mcp add zotero -- uvx zotero-mcp, with the Zotero desktop app running or a ZOTERO_API_KEY), /ideation automatically switches to library-first literature search: your existing papers are recognized and skipped, and new findings can be saved into a project collection. See /setup for guidance.

PubMed / bioRxiv¶

The paper-search-mcp-nodejs server handles both PubMed and bioRxiv searches — no credentials required.

The scholar agent uses it to search NCBI PubMed for peer-reviewed literature and bioRxiv for preprints.

bioRxiv¶

The bioRxiv integration enables the scholar agent to search preprints.

No credentials needed. Works automatically after installation.

Preprints are not peer-reviewed

The scholar agent marks all bioRxiv results with ⚠️ PREPRINT. Preprints have not been peer-reviewed and should be treated with appropriate caution.

bioRxiv API keyword-search limitation

The bioRxiv MCP server uses a date-range API that does not support keyword filtering. When a keyword search returns zero results, the scholar agent will warn you and automatically fall back to CrossRef and Semantic Scholar — both free public APIs that support full keyword search across preprints and peer-reviewed literature. No additional setup is required for the fallback.

Miro¶

The Miro integration enables Claude to create and edit Miro boards — useful for mind maps, experiment diagrams, and visual collaboration during ideation.

Getting a personal access token¶

Go to https://miro.com/app/settings/user-profile/apps
Click Create new app (or select an existing one)
Under Token, click Create token
Copy the token — it starts with eyJ…

No OAuth support

Miro OAuth browser login is not supported from a terminal subprocess. Use a personal access token instead.

Setup¶

Run the wizard:

/neuroflow:setup

Or set the environment variable directly:

export MIRO_ACCESS_TOKEN="eyJhbGciOiJSUzI1NiJ9..."

Context7¶

The Context7 integration provides Claude with up-to-date documentation for libraries like MNE, nilearn, scikit-learn, and PsychoPy — directly in context.

No credentials needed. Works automatically after installation.

This means when Claude writes an MNE preprocessing script, it can look up the current API instead of relying on training data that may be outdated.

Custom LLM providers¶

neuroflow's /setup command (Step 5) lets you configure an alternative LLM API endpoint for Claude Code — replacing Anthropic's API with any OpenAI-compatible endpoint.

e-INFRA CZ (Czech academic researchers only)¶

Access requirement: e-INFRA CZ LLM API is available to Czech academic researchers with Metacentrum/e-INFRA CZ membership. See https://metavo.metacentrum.cz for eligibility. This service is not available for general international use.

The e-INFRA CZ gateway at https://llm.ai.e-infra.cz provides free access to large open-source models — and it speaks the Anthropic protocol natively, so Claude Code connects directly. No proxy needed.

Native connection (recommended):

CLAUDE_CONFIG_DIR="$HOME/.claude-meta" \
ANTHROPIC_BASE_URL="https://llm.ai.e-infra.cz/v1" \
ANTHROPIC_AUTH_TOKEN="<YOUR_API_KEY>" \
ANTHROPIC_MODEL="agentic" \
ANTHROPIC_SMALL_FAST_MODEL="mini" \
ANTHROPIC_DEFAULT_HAIKU_MODEL="mini" \
ANTHROPIC_DEFAULT_SONNET_MODEL="agentic" \
ANTHROPIC_DEFAULT_OPUS_MODEL="agentic" \
CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1 \
claude

The isolated CLAUDE_CONFIG_DIR lets this run alongside your normal subscription claude in another terminal — zero shared state. Get an API key at chat.ai.e-infra.cz → Settings → Account → API keys.

Models: the gateway list changes over time — fetch it live from https://llm.ai.e-infra.cz/v1/models. As of August 2026 it includes agentic (recommended for Claude Code tool use), coder, thinker, mini (background tasks), kimi-k3, qwen3.5-122b, glm-5, deepseek, and more. Switch mid-session with /model <id>.

Rate limit: 4 parallel requests per key — normal interactive use fits; heavy parallel subagent fan-out will hit 429s.

For full documentation — env var reference, context-window handling, and the legacy proxy appendix — see the neuroflow:setup skill → references/einfra-cc.md.

Other providers¶

Any OpenAI-compatible endpoint works: set base_url and api_key during /setup Step 5. The custom LLM settings are saved to your chosen scope's integrations.json under the custom_llm key.

Credential storage¶

When you run /neuroflow:setup, credentials are saved to ~/.neuroflow/integrations.json (global, all projects on this machine) or .neuroflow/integrations.json (per-project override — takes precedence). The wizard asks which scope to use:

{
  "miro": {
    "MIRO_ACCESS_TOKEN": "eyJ..."
  },
  "custom_llm": {
    "provider": "einfra",
    "base_url": "https://llm.ai.e-infra.cz/v1",
    "api_key": "<stored locally, gitignored>",
    "model": "qwen3.5-122b"
  }
}

Never committed

The per-project .neuroflow/integrations.json is automatically added to .gitignore by neuroflow, and the global ~/.neuroflow/integrations.json lives outside any repository. Your credentials are stored locally only and never committed. Only non-secret settings (never api_key) are ever synced via ~/.neuroflow/flowie/integrations.json.

Reminder behavior¶

neuroflow checks credentials at the point of use, not upfront:

Trigger	Reminder
Mention Miro in any command	If `MIRO_ACCESS_TOKEN` missing → offer to run `/setup` or skip
Mention custom LLM / e-INFRA	If `custom_llm` missing → offer to run `/setup` Step 5

You can always re-run /neuroflow:setup to add or update credentials at any time.