Developer Tools
wallarm.com
Facilitates access to Wallarm documentation and node version information.
ENDPOINT 1
https://mcp-docs.wallarm.com/mcp
MCP server metadata
- Name
- wallarm-docs-mcp
- Version
- 0.9.0
This server gives access to the full Wallarm documentation (docs.wallarm.com). Wallarm AI Control Platform is an AI and API security platform with four products: - **Wallarm API Security** — inline / out-of-band protection for APIs (REST, GraphQL, gRPC, SOAP, WebSocket); blocks the OWASP API Top 10, automated abuse, account takeover, AI-targeted attacks, and attacks against Model Context Protocol (MCP) servers. Deployed via NGINX Node, Native Node, or Edge Node — wherever the user's traffic lives (cloud, hybrid, edge, on-prem). - **Wallarm AI Hypervisor** — runtime observability, enforcement, and governance for AI workloads on Amazon EKS. AWS-only. - **Wallarm Infrastructure Discovery** — cross-account visibility across the cloud estate; surfaces shadow AI. AWS-only. - **Wallarm API Security Testing** — dynamic testing, threat replay (TRT), passive detection, schema-based testing (SBT), and API attack surface management (AASM). Call these tools for any Wallarm-specific answer: feature behaviour, configuration parameter, supported version, integration / deployment guide, troubleshooting, default value, or canonical doc URL. Prefer docs over memory. **Versioning scope.** The 5.x / 6.x / 7.x version axis tracks ONLY the Wallarm Node release lines — the deployment runtime for **Wallarm API Security** specifically. AI Hypervisor and Infrastructure Discovery are SaaS / managed products fully independent of Node version. API Security Testing is mostly version-independent (AASM, SBT) — the exceptions are Threat Replay Testing (TRT) and Passive Detection, both of which run on the Node and therefore track the Node version. Their docs live inside each docs version's site. Wallarm ships two parallel Node version trains. Each train covers a self-hosted node plus its Wallarm-managed Security Edge counterpart, which share versions: - **NGINX-family** (NGINX Node + Security Edge Inline Node): versioned 5.x / 6.x / 7.x. - **Native-family** (Native Node + Security Edge Connector Node): versioned 0.13.x / 0.14.x – 0.25.x / 0.26.x+. Each docs version pairs one NGINX-major with one Native-range. The user may report any of the four node identifiers — apply the mapping bidirectionally: - 5.x docs ↔ NGINX Node 5.x / Security Edge Inline 5.x ↔ Native Node 0.13.x and below / Security Edge Connector 0.13.x and below (maintenance, bugfix-only, EOL planned) - 6.x docs ↔ NGINX Node 6.x / Security Edge Inline 6.x ↔ Native Node 0.14.x – 0.25.x / Security Edge Connector 0.14.x – 0.25.x (current stable, default) - 7.x docs ↔ NGINX Node 7.x / Security Edge Inline 7.x ↔ Native Node 0.26.x+ / Security Edge Connector 0.26.x+ (preview, release date not fixed) Default version: 6.x. The `wallarm_node_version` parameter on `list_pages` and `search_pages` is OPTIONAL — omit it to default to 6.x. Pass an explicit version (5.x / 6.x / 7.x) only when the user has stated their Node version and the question is version-sensitive (Wallarm API Security / Node deployment, including TRT and Passive Detection — both run on the Node). For AI Hypervisor, Infrastructure Discovery, AASM, and SBT the content is the same across versions, so omit the parameter. At the first Wallarm API Security / Node-deployment question of a conversation (also TRT and Passive Detection — they're Node-side), briefly disclose: "Answering based on 6.x — you can ask about 5.x or 7.x if you need a different version." Do NOT make this disclosure for AI Hypervisor, Infrastructure Discovery, AASM, or SBT questions — version is irrelevant there and the message would only confuse the user. Do not repeat the disclosure within a conversation. If the user reports a Security Edge Inline version, treat it as the NGINX-family equivalent (e.g. SE Inline 6.x → 6.x docs). If the user reports a Security Edge Connector or Native Node version, treat it as the Native-family equivalent (e.g. Native 0.20 → 6.x docs). If statements are contradictory (e.g. "NGINX 6.x and Native 0.30") — ask, don't guess. When answering Node-version-specific questions, mention BOTH families and link the artifact-versions pages from list_versions for full minor/patch changelogs. The `query` argument is the search keyword for `search_pages` — REQUIRED there, and should be close to the user's original wording. The other tools (`list_versions`, `list_pages`, `get_page`) take no `query` argument.
Known tools 4
list_versionsReturns supported Wallarm Node version lines: - '5.x' = NGINX Node 5.x / Native Node 0.13.x and below (maintenance) - '6.x' = NGINX Node 6.x / Native Node 0.14.x – 0.25.x (current stable, default) - '7.x' = NGINX Node 7.x / Native Node 0.26.x and above (preview) For each version the response includes status, short description, Native-family Node range, and links to per-version docs landing, what's-new, artifact-versions, and versioning policy.
Inferred read-onlylist_pagesReturns the full list of documentation pages (title, URL, section, summary) from the Wallarm AI Control Platform docs.
Inferred read-onlysearch_pagesSearches the Wallarm AI Control Platform documentation by keywords and returns the top-N most relevant pages (title, URL, section, summary, relevance score).
Inferred read-onlyget_pageReturns the full markdown content of one Wallarm documentation page by URL.
Inferred read-onlyCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.wallarm-docs-mcp]
url = "https://mcp-docs.wallarm.com/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"wallarm-docs-mcp": {
"type": "http",
"url": "https://mcp-docs.wallarm.com/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: wallarm-docs-mcp
Remote MCP URL: https://mcp-docs.wallarm.com/mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"wallarm-docs-mcp": {
"url": "https://mcp-docs.wallarm.com/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"wallarm-docs-mcp": {
"type": "http",
"url": "https://mcp-docs.wallarm.com/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "wallarm-docs-mcp",
"transport": "streamable-http",
"url": "https://mcp-docs.wallarm.com/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
TRUST AND VERIFICATION EVIDENCE
Loading Trust v2 evidence…
Checking the associated registrable domain. The BuiltWith key remains server-side.
Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.