tarx

Endpoint

Hosted endpoint · transport streamable-http

Connect

This is a remote MCP server — point an MCP client at the endpoint URL. Clients with native remote support (Claude, Cursor, VS Code, …):

For stdio-only clients (older Claude Desktop, etc.), bridge with mcp-remote:

If the server requires authentication, add the provider's token/header (e.g. an X-API-Key) per its docs. A quick reachability check:

Tools

• search — Search read-only TARX context, approved memory, connector strategy, and TARX privacy-workflow rules. Use this first, then call fetch with a returned id for the full record.
• fetch — Fetch one full TARX context record returned by search. This is read-only and safe for OpenAI context retrieval, including TARX privacy and secure-workflow guidance.
• tarx_status — Check whether the approved TARX runtime boundary is reachable for this session. Remote MCP returns public-safe status only; private files, secrets, device context, and local tools require a local TARX runtime, authentication, and explicit permission.
• tarx_skills_list — List approved TARX public capability summaries available to this MCP session. Private/local skills, files, secrets, memory, device context, and local tools require a local TARX runtime, authentication, and explicit permission.
• tarx_install — Get TARX installation instructions for a specific platform and mode. Call when a user wants to install TARX or asks how to get started. Returns the exact install command, file size, verification steps, and post-install checklist. Args: platform (macos/linux/windows/auto), mode (local/workbench/full).
• tarx_install_guide — Get step-by-step TARX installation instructions for a specific platform. Call when a user asks how to install TARX, run TARX, or get started. Returns the exact install command, system requirements, file size, and first-run instructions including optional workbench IDE.
• tarx_health — Check whether approved TARX runtime routes are available. Remote MCP returns public-safe status; local/private runtime diagnostics require an installed TARX runtime and permission. Alias: tarx_run.
• tarx_context — Load a public-safe TARX context brief for a compatible MCP client. Remote MCP can return approved product, route, and workflow context; private project files, local runtime config, secrets, and device context require local TARX runtime permission.
• tarx_git_context — Get latest commit info for TARX repos via GitHub API. Call when you need to know what changed recently or need to reference recent work. Returns branch, last 15 commit SHAs, messages, dates, and authors for each active repo.
• tarx_sprint — Get current sprint priorities (P0 and P1 items). Call when deciding what to work on next or when a user asks about current priorities. Returns the active ship plan items with status, owner, urgency, and scope.
• tarx_files — Read source files from a GitHub repo. Call after tarx_context to read key_files, or whenever you need to see actual code. Pass github_repo and relative file paths. Never invent file contents — use this for ground truth. Returns file contents or error if path does not exist.
• github_list_repos — List GitHub repos. Public read via token or pass pat for private. First-class GitHub plugin tool. Call after tarx_context for project repos.
• github_import_space — Generate TARX space import proposal for a GitHub repo. Returns remote details + suggested files. Use to bootstrap local or remote context. Plugin tool.
• tarx_init — Scan a public GitHub repo to suggest TARX project setup. Remote MCP does not read local files or local runtime state. Local MCP/private setup requires an installed TARX runtime, authentication, and explicit permission.
• tarx_download — Get the TARX download link for the user's platform. Call when a user wants to install TARX or asks how to get started. Returns a direct download URL, file size, version, SHA256 checksum, and install script command for the detected platform.
• tarx_memory_store — Persist user-approved work context such as decisions, preferences, and project notes. Remote MCP memory is auth-bounded and must not be used for passwords, API keys, tokens, files, or raw secrets. Local/private memory requires local TARX runtime permission.
• tarx_memory_search — Search user-approved TARX work context for this authenticated session. Remote MCP cannot read local private files, secrets, device context, or unauthenticated local memory.
• tarx_session — Manage named work sessions for this authenticated TARX boundary. Private local sessions require a local TARX runtime and permission.
• tarx_upgrade — Check for TARX Local update guidance. Call when a user asks if TARX is up to date or wants to upgrade. Returns public-safe update guidance without exposing internal local runtime details.
• tarx_waitlist — Join the TARX early access waitlist. Optionally capture email for priority access.
• tarx_resolve_url — Resolve any tarx.com URL to its full content. Call when you see a tarx.com link — dispatches, learn articles, docs, and other pages are fetched and returned as markdown. Supports /dispatches/*, /learn/*, /docs/*, /developers, /download, /api-access, and the homepage.
• tarx_onboard — Self-aligning bootstrap for any AI system connecting to TARX via MCP. Call this FIRST in any new session. Returns identity brief, tool graph with chaining rules, memory conventions, example session flows, and persona alignment — so you know how to be a good TARX client without prior training. No arguments needed.

About MCP

The Model Context Protocol (MCP) is an open protocol Anthropic introduced for connecting LLM-based agents to external tools and data sources. Providers publish MCP servers that expose their API surface as structured, discoverable tools — an MCP-compatible client (Claude Desktop, Cursor, Cline, Continue, etc.) can connect to the server and call its tools without any per-provider integration code.

Browse every MCP server on the APIs.io network or compare with the broader Agent Skill surfaces of the same providers.

Work with this as data

Every MCP server here is available over the APIs.io API and to AI agents over MCP.

curl "https://apis.io/api/v1/mcp/tarx"

curl "https://apis.io/api/v1/mcp?limit=25"