onboardingWelcome guide for new users. Returns setup instructions, available tools, pricing, and links. No auth required. Call this when someone is new to CI-1T
evaluateEvaluate prediction stability. Sends scores to the CI-1T engine and returns per-episode stability metrics. Accepts floats (0.0–1.0) or Q0.16 integers
fleet_evaluateEvaluate a fleet of model nodes for prediction stability. Each node provides a score stream. Returns per-node episodes and aggregate fleet stats. Acce
probeProbe an LLM for prediction instability. Sends the same prompt 3 times and compares responses using the specified similarity method. Two modes: (1) De
healthCheck CI-1T engine health. Response: { status, version, latency_ms }. Call before evaluate/fleet operations to verify the engine is reachable.
fleet_session_createCreate a new persistent fleet monitoring session. Returns a session ID for subsequent rounds. Max 16 nodes. Response: { session_id, node_count, node_n
fleet_session_roundSubmit a scoring round to an existing fleet session. Each node
fleet_session_stateGet the current state of a fleet session without submitting new scores. Response: { session_id, round_count, nodes: [{ node_id, node_name, episodes: [
fleet_session_listList all active fleet sessions. Response: { sessions: [{ session_id, node_count, round_count, created_at }] }.
fleet_session_deleteDelete a fleet session by ID. Response: { deleted: true, session_id }. Call when monitoring is complete to free server resources.
list_api_keysList all API keys for the authenticated user. Response: { keys: [{ id, name, masked_key, scope, enabled, created }] }. Use the record id with delete_a
create_api_keyCreate a new CI-1T API key. Response: { api_key, masked_key, scope, record }. IMPORTANT: Save the returned api_key — it cannot be retrieved again afte
delete_api_keyDelete an API key by its PocketBase record ID (from list_api_keys). Response: { deleted: true }.
get_invoicesGet billing history (Stripe invoices). Response: { invoices: [{ amount, credits, date, status }], has_more, cursor }. Pass cursor to paginate.
interpret_scoresAnalyze raw prediction scores with statistical breakdown — no API call, no auth, no credits. Response: { count, mean, std, min, max, breakdown: [{ ind
convert_scoresConvert between probability floats (0.0–1.0) and Q0.16 fixed-point integers (0–65535) — no API call, no auth, no credits. Response: { direction, count
generate_configGenerate CI-1T integration boilerplate for a specific framework or language — no API call, no auth, no credits. Response: { framework, use_case, api_b
compare_windowsCompare two windows of episodes to detect drift or degradation — no API call, no auth, no credits. Takes baseline and recent episode arrays from evalu
+2 more tools
We re-grade npm:@collapseindex/ci1t-mcp on a schedule and alert your Slack/webhook the moment its tools change or its grade drops — rug-pull insurance for the connection.
Add the wmcp.sh trust oracle as an MCP server and call grade_mcp_server / check_mcp_drift in your agent's pre-connection gate:
https://wmcp.sh/mcp/trust
readOnly vs observed behavior) layer on via the wmcp.sh proxy.