# VaultCrux Platform (remote · api.vaultcrux.com)

VaultCrux Platform — 60 tools: retrieval, proof, intel, economy, watch, org

- Trust score: 25/100 (low)
- Change this week: −28
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- remote · `api.vaultcrux.com`: 25/100 (this document), [markdown](https://verifymcp.io/servers/cuecrux-vaultcrux-platform/platform-mcp.md), [page](https://verifymcp.io/servers/cuecrux-vaultcrux-platform/platform-mcp)

## Channel facts

- Endpoint: `https://api.vaultcrux.com/platform/mcp`
- Transports: `streamable-http`
- Auth: `required`
- Version: `0.1.0`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Endpoint Security**: 63/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - Authorisation not fully verified: no authorisation is required to connect, but we couldn't read the tool list to see what that exposes.
  - HTTPS is enforced; there's no plaintext access path.
  - The HSTS (Strict-Transport-Security) header is present.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 0/100
  - Transport check failed: declared streamable-http, but the endpoint returned HTTP 502.
- **Schema Quality & AI Usability**: 0/100
  - Schema not yet verified: we couldn't read the endpoint's schema.
- **Stability & Change Management**: 0/100
  - Stability not yet verified: not enough scan history yet (needs a 30-day window).
- **Tool Coverage**: 0/100
  - Tool coverage not yet verified: we couldn't read the endpoint's tools.
- **Capabilities**: 0/100
  - Capabilities not yet verified: we couldn't read the endpoint's capabilities.

**Unverified: 4 categories.** Categories scored 0 because we could not verify them: authentication we do not have, an unreachable endpoint, or not enough scan history. We only credit what we can confirm.

## Install

### Claude

```bash
claude mcp add --transport http cuecrux-vaultcrux-platform https://api.vaultcrux.com/platform/mcp
```

### Codex

```toml
[mcp_servers.cuecrux-vaultcrux-platform]
url = "https://api.vaultcrux.com/platform/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "cuecrux-vaultcrux-platform": {
      "type": "remote",
      "url": "https://api.vaultcrux.com/platform/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add cuecrux-vaultcrux-platform --url https://api.vaultcrux.com/platform/mcp --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  cuecrux-vaultcrux-platform:
    url: "https://api.vaultcrux.com/platform/mcp"
```

### Other

```json
{
  "mcpServers": {
    "cuecrux-vaultcrux-platform": {
      "type": "http",
      "url": "https://api.vaultcrux.com/platform/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-07-31 (score 25, +9)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-28 (score 16, −37)

- [security regression] Endpoint reachability: reachable → not serving MCP
- [security regression] Stability: 0.03 → unverified
- [security regression] Transport: pass → fail
- [security] Authorization: Authorisation not fully verified: no authorisation is required to connect, but we couldn't read the tool list to see what that exposes.
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional] First check of Schema quality: unverified

### 2026-07-27 (score 53, +1)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-26 (score 52)

First indexed and scored.

## MCP tools (76)

### `cuecrux_session` (~226 tokens)

CueCrux Session

Opens a CueCrux session and returns a typed capability plan (retrieval, proofing, memory, journaling, audit) across VaultCrux and MemoryCrux. Call this first, once. Every subsequent action routes through the channels the plan returns — do not browse the legacy per-service tool list when a plan channel already covers the capability. Identical behaviour for local Crux CE and hosted CueCrux. Hosted deployments stage v1 flat-list or v2 typed-graph plan shapes behind feature flags; callers treat the returned plan as the single source of routing truth. Bulk-capable agents transparently use the HTTP/2 binary channel; MCP-only agents use the MCP fallback URLs the plan provides. Implements RCX-Protocol v1.0.

Input parameters:

- `hints` (object): Optional routing / shaping hints.
- `intent` (string): Optional intent hint (e.g., 'audit_review', 'document_ingest'). Lets the capability graph be reordered to put intent-relevant capabilities first.
- `model` (object): Optional model declaration used for capability-model policy gating.

### `query_vault` (~225 tokens)

Query Vault

Retrieve relevant documents from the vault using semantic search across one or more corpora.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `corpusIds` (array): Corpus IDs to search within.
- `includeCommons` (boolean): Whether to include common/shared corpora in the search.
- `lane` (string): Retrieval lane controlling depth and cost. Accepts `light|verified|audit` — map informal terms (e.g. 'quick'→`light`, 'strict'→`audit`) to the nearest enum value. Defaults to `light`.
- `limit` (integer): Maximum number of results to return (1-50, default 8).
- `query` (string, required): The search query to retrieve documents for.

### `query_with_threshold` (~165 tokens)

Query with Trust Threshold

Execute a trust-routed query that filters results by minimum confidence and respects budget constraints.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `budget_cap` (number): Maximum budget units to spend on this query.
- `min_confidence` (number): Minimum confidence threshold (0-1, default 0.8).
- `query` (string, required): The search query.
- `requested_mode` (string): Requested routing mode override.

### `get_session_context` (~104 tokens)

Get Session Context

Retrieve the current session context for the active agent, including recent interactions and state.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `get_journal` (~163 tokens)

Get Journal

Fetch journal entries for the active agent, with optional filtering by time range, entry type, and pagination.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cursor` (string): Pagination cursor from a previous response.
- `limit` (integer): Maximum number of entries to return.
- `since` (string): ISO 8601 datetime to fetch entries after.
- `types` (array): Entry types to filter by.

### `action_journal_query` (~200 tokens)

Query Action Journal Receipts

Search action journal receipts with filters for time range, outcome, and tool name.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `limit` (integer): Maximum number of receipts to return.
- `offset` (integer): Number of receipts to skip for pagination.
- `outcome` (string): Filter by outcome status.
- `since` (string): ISO 8601 datetime for the start of the query window.
- `tool_name` (string): Filter by the name of the tool that produced the receipt.
- `until` (string): ISO 8601 datetime for the end of the query window.

### `get_stale_pins` (~134 tokens)

Get Stale Pins

List pinned items for the active agent that may be outdated and need refresh or removal.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cursor` (string): Pagination cursor from a previous response.
- `limit` (integer): Maximum number of stale pins to return.

### `memory_retrieve` (~314 tokens)

Memory Retrieve

Retrieve memory chunks, optional curated ESI facts, and passport-driven engrams for Pattern B memory reasoning. The pre_logic field in the response is a ready-to-inject system prompt preamble containing structural data-shape facts calibrated to the calling model's capability class — insert it before reasoning. When deterministic engram pre-execution is enabled, the server may also inline enumerated_facts directly in the response.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `groupId` (string): Optional enrichment config group.
- `iteration` (integer): 1-based retrieval iteration number.
- `modelId` (string): The LLM model ID making this call (e.g. 'claude-sonnet-4-6'). Used to calibrate which engrams are dispatched and how pre_logic is formatted. Omit if unknown.
- `query` (string, required): The memory question or retrieval query.
- `sessionId` (string): Optional session identifier for receipt grouping.
- `sessionProcedureHash` (string): Hash of a session_procedure already seen by this client. When current, the server omits the procedure body.
- `topicHints` (array): Optional topic hints.

### `memory_session_init` (~163 tokens)

Memory Session Init

Boot a memory reasoning session and return the server-controlled session procedure, deterministic passport identity, capability class, and engram manifest. Call once before memory_retrieve when the session-init endpoint is enabled.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `modelId` (string): The LLM model ID making this call (e.g. 'claude-sonnet-4-6'). Used to calibrate the manifest.

### `memory_engrams_resolve` (~169 tokens)

Memory Engrams Resolve

Resolve body content for engrams listed in memory_session_init.available_on_demand. Use when a question matches an on-demand engram's applies_when or trigger_features.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `manifestHash` (string): Manifest hash from memory_session_init.
- `modelId` (string): The LLM model ID making this call.
- `names` (array, required): Engram names in name@version form.

### `memory_reason_about` (~181 tokens)

Memory Reason About

Reason over previously retrieved memory chunks and optional curated facts using the cached Pattern B prompt.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `chunks` (array, required): Chunks returned by one or more memory_retrieve calls.
- `facts` (array): Optional ESI facts returned by memory_retrieve.
- `intent` (string, required): Intent returned by memory_retrieve.
- `question` (string, required): The user question to answer.
- `retrievalReceiptIds` (array, required): Receipt IDs from prior memory_retrieve calls.

### `proof_document` (~173 tokens)

Proof Document

Submit a document artefact for cryptographic proof. Creates an async proof job that retrieves the artefact, chunks it, hashes each chunk, and produces a Merkle receipt. Returns the job ID for status polling.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `artefact_id` (string, required): The artefact ID to proof
- `metadata` (object): Optional metadata to attach to the proof job
- `mode` (string): Proof mode (default: light)

### `get_proof_status` (~129 tokens)

Get Proof Status

Poll the status of a proof job. Returns the current state (queued, processing, complete, failed) and progress details.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `proof_job_id` (string, required): The proof job ID to check

### `get_proof_chunks` (~133 tokens)

Get Proof Chunks

Retrieve the chunk-level hashes for a completed proof job. Supports cursor-based pagination.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cursor` (string): Pagination cursor for the next page of chunks
- `proof_job_id` (string, required): The proof job ID

### `get_proof_receipt` (~130 tokens)

Get Proof Receipt

Retrieve the cryptographic proof receipt for a specific answer. Contains the Merkle root, signature, and verification metadata.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_id` (string, required): The answer ID to get the proof receipt for

### `get_proofpack` (~131 tokens)

Get Proofpack

Download the full proofpack bundle for a receipt. Includes all chunk hashes, Merkle tree, signature, and verification instructions.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `receipt_id` (string, required): The receipt ID to get the proofpack for

### `get_credit_balance` (~104 tokens)

Get Credit Balance

Retrieve the current credit balance for the agent, including receipt verification and passport data.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `get_credit_escrow` (~98 tokens)

Get Credit Escrow

List active escrow holds for the tenant.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `get_spend_receipt` (~112 tokens)

Get Spend Receipt

Retrieve a specific spend receipt by ID.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `receipt_id` (string, required): The receipt ID to look up.

### `get_pricing` (~96 tokens)

Get Pricing

Retrieve current pricing information for the tenant.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `get_economy_dashboard` (~108 tokens)

Get Economy Dashboard

Retrieve the economy dashboard for the agent, showing balances, recent transactions, and spending summaries.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `tip_platform` (~129 tokens)

Tip Platform

Send a credit tip to the platform. Amount must be a positive number.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `amount` (number, required): Tip amount (must be > 0).
- `reason` (string): Optional reason for the tip.

### `browse_bundles` (~109 tokens)

Browse Bundles

List available credit bundles for purchase.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `limit` (integer): Maximum number of bundles to return.

### `purchase_bundle` (~119 tokens)

Purchase Bundle

Purchase a credit bundle by ID.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `bundle_id` (string, required): The bundle ID to purchase.
- `metadata` (object): Optional metadata for the purchase.

### `get_passport` (~95 tokens)

Get Passport

Retrieve the agent's trust passport.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `verify_passport` (~113 tokens)

Verify Passport

Verify another agent's trust passport.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `principal_id` (string): Principal ID to verify (defaults to own agent ID).

### `get_trust_level` (~98 tokens)

Get Trust Level

Get the current agent's trust escalation level.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `register_belief` (~165 tokens)

Register Belief

Register a belief about an answer for trust tracking.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_id` (string): The answer ID this belief relates to.
- `confidence_band` (object): Confidence band object.
- `cost_crux` (number): Credit cost of the belief.
- `decision_context` (string): Context for the decision.
- `receipt_id` (string): The receipt ID backing this belief.

### `get_beliefs` (~145 tokens)

Get Beliefs

List registered beliefs with optional filters.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_ids` (array): Filter by answer IDs.
- `limit` (integer): Maximum number of results.
- `offset` (integer): Offset for pagination.
- `since` (string): ISO 8601 timestamp to filter beliefs after.

### `diff_receipts` (~120 tokens)

Diff Receipts

Compare two provenance receipts and highlight differences.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `receipt_id_a` (string, required): First receipt ID.
- `receipt_id_b` (string, required): Second receipt ID.

### `get_blast_radius` (~141 tokens)

Get Blast Radius

Estimate the impact radius if an artefact or receipt changes.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `artefact_id` (string): The artefact ID to analyze.
- `domain` (string): Domain scope for the analysis.
- `receipt_id` (string): The receipt ID to analyze.

### `get_break_analysis` (~140 tokens)

Get Break Analysis

Analyze what would break if a given answer or receipt is invalidated.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_id` (string): The answer ID to analyze.
- `domain` (string): Domain scope for the analysis.
- `receipt_id` (string): The receipt ID to analyze.

### `get_counterfactual_summary` (~126 tokens)

Get Counterfactual Summary

Generate a counterfactual summary for an answer.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_id` (string): The answer ID to summarize.
- `receipt_id` (string): The receipt ID to summarize.

### `explain_last_answer` (~114 tokens)

Explain Last Answer

Explain how the last answer was derived.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_id` (string): The answer ID to explain (defaults to last answer).

### `get_domain_affinity` (~97 tokens)

Get Domain Affinity

Get the agent's domain affinity scores.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `annotate_session` (~130 tokens)

Annotate Session

Add an annotation to the current session.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `annotation_type` (string): Type of annotation.
- `content` (string, required): Annotation content.
- `session_id` (string): Session ID (defaults to 'default').

### `get_knowledge_gaps` (~127 tokens)

Get Knowledge Gaps

Identify knowledge gaps across domains.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `domain` (string): Filter by domain.
- `limit` (integer): Maximum number of results.
- `offset` (integer): Offset for pagination.

### `get_daily_briefing` (~96 tokens)

Get Daily Briefing

Get the daily knowledge briefing.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `set_policy` (~137 tokens)

Set Policy

Set or update an active policy for the agent.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `policy_name` (string): Policy name (defaults to 'default').
- `principal_id` (string): Principal ID to apply the policy to.
- `rules` (object): Policy rules object.

### `get_active_policy` (~106 tokens)

Get Active Policy

Get the currently active policy.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `policy_name` (string): Policy name to retrieve.

### `schedule_recheck` (~148 tokens)

Schedule Recheck

Schedule a periodic re-check of knowledge freshness.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cron_expr` (string): Cron expression (defaults to '0 0 * * *').
- `next_run_at` (string): ISO 8601 timestamp for the next run.
- `scope` (object): Scope object for the recheck.

### `set_reasoning_profile` (~119 tokens)

Set Reasoning Profile

Set the agent's reasoning methodology profile.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `constraints` (array): Reasoning constraints array.
- `methodology` (object): Reasoning methodology object.

### `get_reasoning_profile` (~97 tokens)

Get Reasoning Profile

Get the agent's current reasoning profile.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `pin_receipt` (~137 tokens)

Pin Receipt

Pin a receipt to prevent it from being garbage collected.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `expires_at` (string): ISO 8601 expiry timestamp.
- `reason` (string): Reason for pinning.
- `receipt_id` (string, required): The receipt ID to pin.

### `tip_agent` (~137 tokens)

Tip Agent

Send a credit tip to another agent.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `amount_crux` (number, required): Tip amount in crux credits.
- `reason` (string): Reason for the tip.
- `recipient_principal_id` (string, required): The recipient agent's principal ID.

### `create_coalition` (~171 tokens)

Create Coalition

Create a multi-agent coalition to address a knowledge gap.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `budget_cap_crux` (number): Budget cap in crux credits (defaults to 10).
- `expires_at` (string): ISO 8601 expiry timestamp.
- `gap_description` (string, required): Description of the knowledge gap.
- `initial_pledge_crux` (number): Initial pledge in crux credits (defaults to 1).

### `join_coalition` (~132 tokens)

Join Coalition

Join an existing coalition with a credit pledge.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `coalition_id` (string, required): The coalition ID to join.
- `pledge_crux` (number): Pledge amount in crux credits (defaults to 1).

### `create_handoff_package` (~128 tokens)

Create Handoff Package

Create a handoff package for multi-agent session transfer.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `scope` (object): Scope object for the handoff.
- `session_id` (string): Session ID (defaults to agent ID).

### `accept_handoff_package` (~111 tokens)

Accept Handoff Package

Accept an incoming handoff package.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `package_id` (string, required): The handoff package ID to accept.

### `find_contradictions` (~128 tokens)

Find Contradictions

Scan for contradictions across knowledge sources.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `depth` (string): Scan depth (e.g. 'shallow', 'deep').
- `scope` (object): Scope object to narrow the scan.

### `forecast_obsolescence` (~123 tokens)

Forecast Obsolescence

Forecast which artefacts are likely to become obsolete.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `artefacts` (array): Artefacts to evaluate.
- `domain` (string): Domain to scope the forecast.

### `watch_answer` (~135 tokens)

Watch Answer

Create a watch on an answer to receive alerts when it changes or becomes stale.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `answer_id` (string, required): The ID of the answer to watch.
- `frequency` (string): How often to check for changes (default: daily).

### `unwatch_answer` (~112 tokens)

Unwatch Answer

Remove an existing watch by its watch ID.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `watch_id` (string, required): The ID of the watch to remove.

### `get_watches` (~123 tokens)

Get Watches

List all active watches for the current agent.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cursor` (string): Pagination cursor from a previous response.
- `limit` (integer): Maximum number of watches to return.

### `get_watch_alerts` (~164 tokens)

Get Watch Alerts

Retrieve alerts triggered by a specific watch, with optional filtering by time and pagination.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cursor` (string): Pagination cursor from a previous response.
- `limit` (integer): Maximum number of alerts to return.
- `since` (string): ISO 8601 datetime to fetch alerts after.
- `watch_id` (string, required): The ID of the watch to retrieve alerts for.

### `submit_feature_request` (~158 tokens)

Submit Feature Request

Submit a new feature request or suggestion to the VaultCrux product team.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `category` (string): Category for the request (default: other).
- `description` (string, required): Detailed description of the requested feature.
- `metadata` (object): Additional metadata to attach to the request.
- `title` (string, required): Short title for the feature request.

### `vote_feature_request` (~132 tokens)

Vote on Feature Request

Cast an upvote on an existing feature request to signal interest.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `metadata` (object): Additional metadata to attach to the vote.
- `request_id` (string, required): The ID of the feature request to vote on.

### `get_feature_requests` (~149 tokens)

Get Feature Requests

List feature requests with optional filtering by category and status.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `category` (string): Filter by feature request category.
- `cursor` (string): Pagination cursor from a previous response.
- `limit` (integer): Maximum number of requests to return.
- `status` (string): Filter by feature request status.

### `declare_revenue_willingness` (~219 tokens)

Declare Revenue Willingness

Declare willingness to pay for a feature or category, helping prioritize the product roadmap.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `billing_cycle` (string): Preferred billing cycle (default: monthly).
- `category` (string): Category of the declaration (default: other).
- `confidence` (string): Confidence level in the willingness declaration (default: medium).
- `metadata` (object): Additional metadata to attach.
- `notes` (string): Free-form notes about the declaration.
- `request_id` (string): The ID of a specific feature request this declaration relates to.
- `willingness_band` (string): Price band the agent is willing to pay (default: lt_100).

### `list_seats` (~125 tokens)

List Seats

List all seats (members) in the current organisation.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `cursor` (string): Pagination cursor from a previous response.
- `limit` (integer): Maximum number of seats to return.

### `invite_seat` (~131 tokens)

Invite Seat

Invite a new member to the organisation by email address.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `email` (string, required): Email address of the person to invite.
- `role` (string): Role to assign to the new member (default: member).

### `change_seat_role` (~126 tokens)

Change Seat Role

Change the role assigned to an existing organisation seat.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `role` (string, required): The new role to assign.
- `seat_id` (string, required): The ID of the seat to update.

### `revoke_seat` (~115 tokens)

Revoke Seat

Remove a member from the organisation by revoking their seat.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `seat_id` (string, required): The ID of the seat to revoke.

### `register_agent` (~173 tokens)

Register Agent

Self-register a new agent with the VaultCrux platform. No API key or tenant ID required.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `agent_display_name` (string): A human-readable display name for the agent.
- `agent_framework` (string): The agent framework being used (default: unknown).
- `callback_url` (string): URL for the platform to send callbacks to.
- `framework_fingerprint` (string): Unique fingerprint of the agent framework instance.

### `request_sponsor` (~124 tokens)

Request Sponsor

Request a sponsor for the current agent session. Requires a session token (vcrx_self_ prefixed).

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `session_token` (string, required): Session token obtained from agent registration.

### `list_projects` (~109 tokens)

List Crux Projects

List all projects on the configured Crux Daemon. Mirrors `corecruxd /v1/projects`.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

### `get_project_context` (~111 tokens)

Get Crux Project Context

Detail for a single project — planning target, allowed passports, working tenants.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `project_id` (string, required)

### `list_work` (~149 tokens)

List Crux Work Items

List work items, optionally filtered by project_id, state, tenant_id, assignee_passport. The kanban surface for cross-agent coordination.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `assignee_passport` (string)
- `project_id` (string)
- `state` (string)
- `tenant_id` (string)

### `create_work` (~175 tokens)

Create Crux Work Item

Create a new work item under a project (defaults: state=planned).

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `assignee_passport` (string)
- `body` (string)
- `created_by_passport` (string, required)
- `linked_issue` (string)
- `linked_pr` (string)
- `project_id` (string, required)
- `state` (string)
- `tenant_id` (string)
- `title` (string, required)

### `update_work_state` (~149 tokens)

Update Crux Work State

Move a work item to a new state. If the calling passport has agent_work_gate=true the request queues for human approval (returns applied:false).

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `blocker_reason` (string)
- `by_passport` (string, required)
- `state` (string, required)
- `work_id` (string, required)

### `comment_on_work` (~129 tokens)

Comment on Crux Work Item

Post a comment on a work item — leave context for the next agent or human.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `author_passport` (string, required)
- `body` (string, required)
- `work_id` (string, required)

### `github_search` (~139 tokens)

Search Crux-Indexed GitHub Corpus

Search the indexed GitHub corpus (commits, PRs, issues, comments) under repos selected via /v1/integrations/github/repos.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `query` (string, required)
- `repo` (string)
- `top_k` (integer)

### `github_recent_commits` (~111 tokens)

Recent GitHub Commits (Crux-Indexed)

Recent indexed commits for a selected repo.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `limit` (integer)
- `repo` (string, required)

### `github_open_prs` (~123 tokens)

Open GitHub PRs (Crux-Indexed)

Indexed PRs in a selected repo (filter open client-side via the returned `state` field).

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `limit` (integer)
- `repo` (string, required)

### `github_open_issues` (~126 tokens)

Open GitHub Issues (Crux-Indexed)

Indexed issues in a selected repo. Optional `label` for client-side filtering.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `label` (string)
- `limit` (integer)
- `repo` (string, required)

### `github_comments_since` (~112 tokens)

Recent GitHub Comments (Crux-Indexed)

Recent comments across selected repos — the 'what happened since I last looked' surface.

Prefer `cuecrux_session` as your first and only direct MCP call. It returns a typed capability plan that routes this tool (and every other) to its preferred channel, tier, and cost class. One call per session is enough; the plan is the source of routing truth for all subsequent work. This tool remains directly callable for backward compatibility; the collapsed surface is the intended surface.

Input parameters:

- `limit` (integer)

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/cuecrux-vaultcrux-platform/platform-mcp#diagnostics

## Score history

- 2026-08-03: 25
- 2026-08-02: 25
- 2026-08-01: 25
- 2026-07-31: 25
- 2026-07-30: 16
- 2026-07-29: 16
- 2026-07-28: 16
- 2026-07-27: 53
- 2026-07-26: 52

## Links

- Remote endpoint: https://api.vaultcrux.com/platform/mcp
- Authorisation metadata: https://api.vaultcrux.com/.well-known/oauth-protected-resource/platform/mcp
- Repository: https://github.com/CueCrux/VaultCrux
- Changelog RSS feed: https://verifymcp.io/servers/cuecrux-vaultcrux-platform/platform-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/cuecrux-vaultcrux-platform/platform-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/cuecrux-vaultcrux-platform/platform-mcp