Direct answer

Download the ZIP, extract the spyn3-data folder into any agent’s supported skills directory, and connect SPYN3 through MCP or the REST environment variables. The skill contains no client-specific commands, credentials or hidden setup: it selects whichever SPYN3 interface the runtime already provides.

Portable package / v1.0

Download once. Use across agents.

The ZIP contains the open-format SKILL.md entrypoint plus focused MCP and REST references. The same files are available individually for agents that ingest skills from URLs.

Download skill ZIPView raw SKILL.md →SHA-256 checksum →
spyn3-data/├── SKILL.md└── references/    ├── mcp.md    └── rest.md

What the skill teaches

Interface neutral

Use native SPYN3 MCP tools when present; otherwise use the REST gateway when its environment is already configured.

Search before detail

Discover ranked snippets first, then load a selected document by stable identifier only when the task needs it.

Evidence attached

Preserve document titles, IDs, connectors and limitations instead of turning generated prose into an unsupported fact.

Read-only boundary

The skill does not authorize syncing, writes, connector changes, administration or broader credentials.

Install it without client lock-in

  1. Download and extract spyn3-data-skill.zip.
  2. Move the resulting spyn3-data directory into the skills location supported by your agent or development environment.
  3. Restart or refresh the agent’s skill discovery if its documentation requires it.
  4. Connect the SPYN3 hosted MCP endpoint, or configure SPYN3_API_URL and SPYN3_API_KEY in the runtime environment.
  5. Ask the agent to search or answer from the connected SPYN3 workspace.
The package never contains your key

Credentials remain in the MCP host or runtime environment. Do not paste a Bearer secret into the skill, chat transcript, repository or client-facing output.

How it chooses MCP or REST

If the runtime exposes search_workspace, query_workspace, list_sources or get_document, the skill follows the MCP reference. If those tools are absent but the two SPYN3 environment variables are configured, it follows the REST reference. If neither route is available, it stops and explains what must be connected without asking the user to reveal a secret.

This is intentionally different from a client configuration. The MCP guide explains how a host reaches SPYN3; the skill explains how an agent should behave after that connection exists. The REST guide documents the fallback surface.

Inspect before installing

The package is plain Markdown. Read the entrypoint, its MCP reference and REST reference directly in the browser. Use the published SHA-256 checksum when an installation workflow needs archive verification.

The skill never expands the authority the user or runtime already provided. It makes the retrieval sequence consistent across agents while leaving the requested deliverable, permissions and external actions under the user’s control.

The interface can change. The evidence discipline should not.

Connect the workspace behind the skill

The portable instructions become useful when an agent can reach one real SPYN3 workspace through MCP or REST. Bring a source, a question and the smallest permission set that can answer it.

Request private beta access
About SPYN3

SPYN3 is AI-ready data infrastructure for technical builders. It connects business sources into a workspace-isolated knowledge layer and exposes that layer through REST and a hosted MCP server, with permission-scoped access and cited retrieval. SPYN3 is currently in private beta.