Connect Liminal and your agent gets 26 tools, among them algorithms, apply_form, apply_grade. It answers at https://liminality.physea.ai/mcp over streamable HTTP, and it listed its tools to an anonymous client, so no sign-in was needed for that. Our probes reached it 9 of 10 times since 4 September 2026; the last check, on 17 October 2026, succeeded.
Reduces a request into grounded sub-questions, routes each to a real tool, returns a reusable route.
Description as the publisher wrote it.
| Tool | What it does |
|---|---|
algorithms | Create, validate, release and run pure algorithms over versioned datasets; runs replay byte-identical with zero model calls. action='run' is one solve; other actions are free. Inputs: action, spec, version_id, changes, dataset_id, name |
apply_form | Apply form answers to sharpen a task. Inputs: question, fills |
apply_grade | Apply a grade-form answer to a result. Inputs: result_id, choice, note |
ask_form | Create a short form for missing user-specific facts. Inputs: question, max_depth, verify_divergence |
catalog | Read the capability catalog: action='search' returns every saved candidate for an objective with distinctiveness deltas; action='delta' compares a named entity pair. Inputs: action, objective, entity_a, entity_b |
composio_connect | Create a caller-scoped connection link for a toolkit. Inputs: toolkit, user_id |
compress_context | Compress long context into a task-focused brief. Inputs: text, task, retention |
delegation | Create, resume, or report a bounded subagent packet (action='create'|'resume'|'report'). Inputs: action, delegation_id, handoff_id, task, role, owned_paths |
export_result | Export a caller-owned result in a supported format. Inputs: result_id, format |
get_capabilities | List solve modes and every public capability. |
get_result_detail | Fetch the full caller-owned result behind a lean solve. Inputs: result_id |
get_settings | Read this caller's saved solve and safety settings. Inputs: user_id |
handoff | Create, safely resume, or list caller-scoped cross-session checkpoints (action='create'|'resume'|'list'). Create accepts an optional subject and source_sessions; Liminality assigns the subject number. Resume by handoff_id — 'latest' resumes Inputs: action, task, session_state, evidence, checkpoint, handoff_id |
manage | Compatibility router for retained-context lifecycle actions. Inputs: action, payload |
optimize | Rank bounded one-factor variants for a route against your stated objective; candidates remain unsaved. Inputs: target, objective, budget, top_k |
recover | Find a workaround for a blocked obligation without changing the task: open a case, read candidates with their boundary, choose, retry, verify against the original criteria (action='open'|'candidates'|'choose'|'retry'|'verify'|'add_candidate Inputs: action, case_id, fields, candidate_id, user_change, outcome |
register_asset | Register a caller-owned resource for future grounding. Inputs: name, kind, attributes, user_id |
report_feedback | Report structured feedback on a result. Inputs: rating, helpful, comment, missing, tool, route_key |
report_outcome | Report verified real-world outcome evidence for a result. Inputs: result_id, worked, note, intent, leaf, endpoint |
research | Run a multi-source research pass for a question. Inputs: question, max_depth |
runs | Run lifecycle with run and attempt identity (action='start'|'status'|'list'|'cancel'|'reconcile'|'attempt'|'outcome'): cancel persists, an unknown effect is reconciled before any retry, simulate never executes. Inputs: action, run_id, operation, idempotency_key, mode, task_id |
set_priorities | Save caller-scoped tasks and importance signals. Inputs: items, user_id |
set_settings | Save caller-scoped settings from one dict (depth, optimization, trace, relay, governance); keys and values in get_capabilities. Inputs: settings, user_id |
solve | Decompose, ground, and answer a non-trivial task. Supports saved or per-call depth, safe trace, caller tools, and lean/full output. Inputs: question, context, resources, tool_manifest, depth, max_reduction_depth |
task | Read or version the contract behind a solve (action='get'|'list'|'versions'|'new_version'|'graph'|'create'|'state'); graph also returns what is still unfinished. Inputs: action, task_id, changes, version, fields, limit |
taxis | Return this caller's deterministic ranked priority frame. Inputs: user_id |
Every block below points at https://liminality.physea.ai/mcp. Pick your client and copy.
claude mcp add --transport http liminal https://liminality.physea.ai/mcp
{
"mcpServers": {
"liminal": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://liminality.physea.ai/mcp"]
}
}
}{
"mcpServers": {
"liminal": {
"url": "https://liminality.physea.ai/mcp"
}
}
}{
"servers": {
"liminal": {
"type": "http",
"url": "https://liminality.physea.ai/mcp"
}
}
}{
"context_servers": {
"liminal": {
"url": "https://liminality.physea.ai/mcp"
}
}
}{
"mcpServers": {
"liminal": {
"serverUrl": "https://liminality.physea.ai/mcp"
}
}
}{
"mcpServers": {
"liminal": {
"url": "https://liminality.physea.ai/mcp"
}
}
}name: Liminal
version: 0.0.1
schema: v1
mcpServers:
- name: Liminal
type: streamable-http
url: https://liminality.physea.ai/mcp{
"mcpServers": {
"liminal": {
"httpUrl": "https://liminality.physea.ai/mcp"
}
}
}{
"mcpServers": {
"liminal": {
"type": "http",
"url": "https://liminality.physea.ai/mcp"
}
}
}{
"mcpServers": {
"liminal": {
"url": "https://liminality.physea.ai/mcp"
}
}
}goose://extension?url=https%3A%2F%2Fliminality.physea.ai%2Fmcp&type=streamable_http&id=liminal&name=Liminal
It answered our last check on 17 October 2026 in 343 ms. Over 10 checks since 4 September 2026 it answered 9. Status here comes from Neuronto's own probes, not from the publisher.
Yes. It is published by physea.ai and served from liminality.physea.ai, the vendor's own domain: https://liminality.physea.ai/mcp.
26, as returned by the server's own tools/list on 8 September 2026: algorithms, apply_form, apply_grade, ask_form, catalog, composio_connect, compress_context, delegation, and 18 more.
It answered our last check on 17 October 2026 in 343 ms. Over 10 checks since 4 September 2026 it answered 9.
Put the server under mcpServers in ~/.cursor/mcp.json with the URL https://liminality.physea.ai/mcp. The exact JSON is above, with the same setup for Claude Code, Claude Desktop, VS Code, Windsurf, Zed, Gemini CLI and the rest.
It listed its tools to an anonymous client, so reading the tool list needs no key. Individual tools can still ask for credentials in their own inputs.
Publisher: ai.physea. Endpoint host: liminality.physea.ai. ARD identifier: urn:air:ai.physea:mcp:liminality.
First indexed 31 August 2026. The tools are exactly what the server returned; nothing on this page
is inferred from its name. Ranking in Neuronto search is never for sale.