Calypso Multimodal RAG MCP

[CALYPSO ADD WEBSITE] Ingests a website URL as bucket-scoped knowledge. One shot: the Calypso backend crawls the URL, generates the title, summary, and tags, persists the website as knowledge, and assigns it to the given buckets in the same call. The normalized URL is the create identity per team: a matching Idempotency-Key replays the existing website, and a collision without one is a typed `website_url_exists` conflict. Requires the `knowledge:website:create` capability (explicit-only) and a backend with POST /v1/knowledge/websites deployed.

[CALYPSO RAG AGENT] Sends each prompt directly to the Calypso agent using the full conversation context. Use this when you want Calypso knowledge retrieval and grounded answers from the RAG backend. Typical requests: - "Summarize the key points from our onboarding documentation" - "What does the knowledge base say about campaign approval rules?" - "Compare the documented indexing flow with the retrieval flow" - "Answer using the uploaded file ids: [\"file_123\"]" Responses API behavior: - First turns start a named Calypso conversation via `/v1/responses`. - Follow-up turns chain with `previous_response_id` so the backend owns conversation state. - When `fileIds` are provided, the MCP uses `rag_policy` retrieval semantics instead of inline attachment stuffing. MCP session behavior: - This tool maintains a stable conversation id in the background for multi-turn retrieval context. - Use `/new` to start a fresh conversation and clear the current context window. Quick commands (examples): - "Summarize the latest indexed knowledge about WhatsApp templates" - "Find the source of truth for campaign approval behavior" - "Start a new topic" (or use `/new`) The authoritative, refreshable variant list lives in the `calypso://rag-agent-models` resource — do not cache this description.

[CALYPSO CREATE AGENT] Creates a Calypso agent variant bound to one or more buckets. The success payload leads with `model` — pass it straight to calypso-agent. Bucket bindings are validated server-side (`bucket_not_found` for unknown or archived buckets), agent ids are slug-normalized (`agent_id_exists` on collision), and the plan's agent cap is enforced (`agent_limit_reached` when full). Requires the `rag:agent:create` capability and a backend with POST /v1/rag-agent/agents deployed.

[CALYPSO CREATE BUCKET] Creates an empty bucket to upload files into later. Use this for the create-then-fill workflow; uploads can also create buckets implicitly via bucketSlugs + createMissingBuckets. The server normalizes the slug and returns 409 `bucket_slug_exists` on collision. Requires the `knowledge:bucket:create` capability on the API key and a backend with POST /v1/knowledge/buckets deployed.

[CALYPSO GET FILE] Fetches one knowledge file's metadata and indexing status by file id. Use this to resolve the opaque `fileIds` returned by calypso-list-buckets into filename, mime type, size, indexing status, and per-bucket sync state. Pass verify=true to also ask the provider (Gemini) for ground truth about the indexed document — slower, but authoritative when status looks stale.

[CALYPSO LIST BUCKETS] Lists buckets for the team tied to the configured Calypso API key. Use this before uploads when you need bucket ids, slugs, names, member counts, or bucket-store readiness. This complements agent model discovery: model discovery shows which buckets are bound to each agent variant, while this tool lists all buckets for the API key team.

[CALYPSO SEARCH] Retrieval-only search over the team's grounded knowledge: returns the chunks and sources an agent answer would cite, without a synthesized answer. Use this when the caller wants raw evidence to reason over. Scope defaults to the default Calypso agent; pass `agent` for a named variant (calypso-agent:{agent_id}) or `buckets` (ids or slugs, up to 5) to search specific buckets directly — not both.

[CALYPSO UPLOAD FILE] Uploads a file into the durable bucket-backed knowledge store and indexing pipeline. Use this when you want a file indexed into the broader knowledge corpus instead of attached directly to a single RAG chat turn. This tool returns file and task metadata. A bucket destination is required: pass bucketIds, bucketSlugs, or bucket. Choose exactly one file source. Use `filePath` when this MCP server runs locally and can read the path, including Claude Desktop or Cursor configs that launch this package with npx. Use `contentBase64` for hosted or remote MCP clients, browser uploads, generated in-memory content, or remote sandbox files that this MCP process cannot read. Use `sourceUrl` for a public http(s) file URL — the Calypso backend fetches it directly (public hosts only, 25MB cap), so the bytes never pass through this MCP process; filename and content type are derived from the remote response. Do not base64-encode local files just to use this tool.

[CALYPSO UPLOAD FILES BATCH] Uploads 1 to 100 files into the durable knowledge store and indexing queue in one request. Use this for bulk corpus ingestion. Shared bucket fields apply to every item unless an item provides its own bucket fields. The tool returns batch-level status and, when requested, polls until the batch reaches active, partially_active, partially_failed, failed, or timeout. A shared bucket destination is required unless every item provides its own bucket destination. Choose exactly one file source per item. Use `filePath` when this MCP server runs locally and can read each path, including Claude Desktop or Cursor configs that launch this package with npx. Use `contentBase64` for hosted or remote MCP clients, browser uploads, generated in-memory content, or remote sandbox files that this MCP process cannot read. Do not base64-encode local files just to use this tool.