Contradiction-free memory, causal root-cause graph & verified Python sandbox for coding agents.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Bearer oxy_live_... API key from https://oxyo.si/app (50 free MCP calls included)
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Inspect the immutable audit trail of superseded contradictions vs current truth.
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http oxyo-mcp https://oxyo.si/mcp
Headers
Authorization (optional, not in the config above): Bearer oxy_live_... API key from https://oxyo.si/app (50 free MCP calls included)Inspect the immutable audit trail of superseded contradictions vs current truth.
| Argument | Type | Description |
|---|---|---|
| limit | integer |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 20
}
}
}Run causal reasoning across Pearl's Ladder: multi-hop `intervene`/`diagnose`, graph-mutilated `do_calculus`, Rung-3 `counterfactual` (Abduction-Action-Prediction), or active `experiments` ranked by Expected Information Gain.
| Argument | Type | Description |
|---|---|---|
| do | object | Optional intervention map {node: 0.0..1.0} for `do_calculus` or `counterfactual` |
| mode | string | |
| scope | string | Optional project/repo scope |
| entity | string | Target or seed causal entity |
| factual | object | Optional observed factual state {node: 0.0..1.0} for `counterfactual` |
| max_hops | integer |
{
"type": "object",
"properties": {
"do": {
"type": "object",
"description": "Optional intervention map {node: 0.0..1.0} for `do_calculus` or `counterfactual`"
},
"mode": {
"enum": [
"intervene",
"diagnose",
"do_calculus",
"counterfactual",
"experiments"
],
"type": "string",
"default": "intervene"
},
"scope": {
"type": "string",
"description": "Optional project/repo scope"
},
"entity": {
"type": "string",
"description": "Target or seed causal entity"
},
"factual": {
"type": "object",
"description": "Optional observed factual state {node: 0.0..1.0} for `counterfactual`"
},
"max_hops": {
"type": "integer",
"default": 4
}
}
}Record or reinforce a directed cause-effect link in Oxyo's causal graph.
| Argument | Type | Description |
|---|---|---|
| cause* | string | |
| scope | string | Optional project/repo scope (default 'global') |
| action | string | |
| effect* | string | |
| polarity | integer | |
| succeeded | boolean |
{
"type": "object",
"required": [
"cause",
"effect"
],
"properties": {
"cause": {
"type": "string"
},
"scope": {
"type": "string",
"description": "Optional project/repo scope (default 'global')"
},
"action": {
"type": "string",
"default": "causes"
},
"effect": {
"type": "string"
},
"polarity": {
"enum": [
1,
-1
],
"type": "integer",
"default": 1
},
"succeeded": {
"type": "boolean",
"default": true
}
}
}Run zero-LLM hippocampal memory consolidation to archive redundant episodes and promote recurring patterns.
| Argument | Type | Description |
|---|---|---|
| scope | string | Optional project/repo scope to consolidate |
| keep_recent | integer |
{
"type": "object",
"properties": {
"scope": {
"type": "string",
"description": "Optional project/repo scope to consolidate"
},
"keep_recent": {
"type": "integer",
"default": 20
}
}
}Manage persistent multi-session Long-Horizon Goal DAG: create goals, decompose sub-goals, query next unblocked actions, and update status with automatic Causal Backtracking on failures.
| Argument | Type | Description |
|---|---|---|
| limit | integer | |
| scope | string | |
| title | string | |
| action | string | |
| status | string | |
| goal_id | string | |
| evidence | string | |
| priority | number | |
| subgoals | array | |
| parent_id | string | |
| depends_on | array | |
| failure_cause | string |
Run exact zero-LLM symbolic constraint solving (`satisfy` + `minimize`/`maximize`) or universal invariant theorem proving (`prove`: verifies `assumptions => claim` or synthesizes a minimal counterexample).
| Argument | Type | Description |
|---|---|---|
| claim | string | Universal invariant expression to prove (for action='prove') |
| sense | string | |
| action | string | |
| objective | string | Optional algebraic objective to optimize (for action='satisfy') |
| variables* | object | Map of variable specs, e.g. {'x': {'type': 'int', 'min': 1, 'max': 20}} |
| assumptions | array | Precondition constraints for action='prove' |
| constraints | array | Symbolic constraints to satisfy (for action='satisfy') |
{
"type": "object",
"required": [
"variables"
],
"properties": {
"claim": {
"type": "string",
"description": "Universal invariant expression to prove (for action='prove')"
},
"sense": {
"enum": [
"minimize",
"maximize"
],
"type": "string",
"default": "minimize"
},
"action": {
"enum": [
"prove",
"satisfy"
],
"type": "string",
"default": "prove"
},
"objective": {
"type": "string",
"description": "Optional algebraic objective to optimize (for action='satisfy')"
},
"variables": {
"type": "object",
"description": "Map of variable specs, e.g. {'x': {'type': 'int', 'min': 1, 'max': 20}}"
},
"assumptions": {
"type": "array",
"items": {
"type": "string"
},
"description": "Precondition constraints for action='prove'"
},
"constraints": {
"type": "array",
"items": {
"type": "string"
},
"description": "Symbolic constraints to satisfy (for action='satisfy')"
}
}
}Search persistent Oxyo memory: surgical episodic snippets, active contradiction-free beliefs (zero-noise filtered), cause-effect chains, and verified Python skills.
| Argument | Type | Description |
|---|---|---|
| query* | string | Semantic and keyword search query |
| scope | string | Optional project/repo scope (inherits 'global' rules automatically) |
| k_beliefs | integer | |
| k_episodes | integer | |
| full_episodes | boolean | |
| reinforce_ids | array | Optional memory IDs that proved useful, reinforcing recall plasticity |
| max_episode_chars | integer |
{
"type": "object",
"required": [
"query"
],
"properties": {
"query": {
"type": "string",
"description": "Semantic and keyword search query"
},
"scope": {
"type": "string",
"description": "Optional project/repo scope (inherits 'global' rules automatically)"
},
"k_beliefs": {
"type": "integer",
"default": 8
},
"k_episodes": {
"type": "integer",
"default": 4
},
"full_episodes": {
"type": "boolean",
"default": false
},
"reinforce_ids": {
"type": "array",
"items": {
"type": "string"
},
"description": "Optional memory IDs that proved useful, reinforcing recall plasticity"
},
"max_episode_chars": {
"type": "integer",
"default": 560
}
}
}Persist a durable fact, rule, or user preference. Deduplicates paraphrases and supersedes contradicted older 1:1 claims while keeping 1:N additive facts intact.
| Argument | Type | Description |
|---|---|---|
| claim* | string | |
| scope | string | Optional project/repo scope (default 'global') |
| confidence | number |
{
"type": "object",
"required": [
"claim"
],
"properties": {
"claim": {
"type": "string"
},
"scope": {
"type": "string",
"description": "Optional project/repo scope (default 'global')"
},
"confidence": {
"type": "number",
"default": 0.88
}
}
}Store an experience, architectural decision, or observation with dynamic neural salience.
| Argument | Type | Description |
|---|---|---|
| kind | string | |
| scope | string | Optional project/repo scope (default 'global') |
| content* | string | |
| importance | number | |
| use_neural_salience | boolean |
{
"type": "object",
"required": [
"content"
],
"properties": {
"kind": {
"enum": [
"observation",
"action",
"thought",
"user"
],
"type": "string",
"default": "observation"
},
"scope": {
"type": "string",
"description": "Optional project/repo scope (default 'global')"
},
"content": {
"type": "string"
},
"importance": {
"type": "number",
"default": 0.65
},
"use_neural_salience": {
"type": "boolean",
"default": true
}
}
}Run autonomous self-play problem verification and curated trace export.
| Argument | Type | Description |
|---|---|---|
| seed | integer | Optional deterministic curriculum seed |
| difficulty | integer | Difficulty tier (1..5) |
| max_problems | integer | Number of self-play problems to verify (1..8) |
| export_corpus | boolean | Export curated verified trace dataset after verification |
| run_self_play | boolean | Generate & verify new parametric problems before curating |
| max_per_family | integer | Maximum traces retained per problem family |
| promote_skills | boolean | Promote reusable verified helper functions to the skill vault |
{
"type": "object",
"properties": {
"seed": {
"type": "integer",
"description": "Optional deterministic curriculum seed"
},
"difficulty": {
"type": "integer",
"default": 2,
"description": "Difficulty tier (1..5)"
},
"max_problems": {
"type": "integer",
"default": 3,
"description": "Number of self-play problems to verify (1..8)"
},
"export_corpus": {
"type": "boolean",
"default": true,
"description": "Export curated verified trace dataset after verification"
},
"run_self_play": {
"type": "boolean",
"default": true,
"description": "Generate & verify new parametric problems before curating"
},
"max_per_family": {
"type": "integer",
"default": 3,
"description": "Maximum traces retained per problem family"
},
"promote_skills": {
"type": "boolean",
"default": true,
"description": "Promote reusable verified helper functions to the skill vault"
}
}
}One-call cold-start session capsule (<350 tokens): returns active contradiction-free project/global beliefs, recent actions, causal bug warnings, and verified skills.
| Argument | Type | Description |
|---|---|---|
| focus | string | Optional current task focus query to rank rules and warnings |
| scope | string | Optional project/repo scope (e.g. 'oxyo', 'omniamus', or 'global') |
| auto_consolidate | boolean | Automatically compact redundant older episodes when above threshold |
{
"type": "object",
"properties": {
"focus": {
"type": "string",
"default": "",
"description": "Optional current task focus query to rank rules and warnings"
},
"scope": {
"type": "string",
"description": "Optional project/repo scope (e.g. 'oxyo', 'omniamus', or 'global')"
},
"auto_consolidate": {
"type": "boolean",
"default": true,
"description": "Automatically compact redundant older episodes when above threshold"
}
}
}List or execute a verified Python skill from private or collective Oxyo library.
| Argument | Type | Description |
|---|---|---|
| args | array | |
| name | string | |
| action | string |
{
"type": "object",
"properties": {
"args": {
"type": "array",
"items": {}
},
"name": {
"type": "string"
},
"action": {
"enum": [
"list",
"execute"
],
"type": "string",
"default": "list"
}
}
}Register a reusable (and optionally compositional) Python function after verifying 100% of its test cases.
| Argument | Type | Description |
|---|---|---|
| code* | string | |
| name* | string | |
| entrypoint* | string | |
| test_cases* | array | |
| description* | string | |
| uses_skills | array | Optional names of existing verified skills called by this skill |
| share_collective | boolean |
{
"type": "object",
"required": [
"name",
"entrypoint",
"description",
"code",
"test_cases"
],
"properties": {
"code": {
"type": "string"
},
"name": {
"type": "string"
},
"entrypoint": {
"type": "string"
},
"test_cases": {
"type": "array",
"items": {
"type": "object"
}
},
"description": {
"type": "string"
},
"uses_skills": {
"type": "array",
"items": {
"type": "string"
},
"description": "Optional names of existing verified skills called by this skill"
},
"share_collective": {
"type": "boolean",
"default": true
}
}
}Consolidate verified execution traces into tested, reusable functions in your persistent skill vault and compact episodic memory.
| Argument | Type | Description |
|---|---|---|
| scope | string | |
| traces | array | Optional inline verified trace objects [{'prompt': ..., 'verified_code': ...}] |
| keep_recent | integer | |
| max_new_skills | integer |
{
"type": "object",
"properties": {
"scope": {
"type": "string"
},
"traces": {
"type": "array",
"items": {
"type": "object"
},
"description": "Optional inline verified trace objects [{'prompt': ..., 'verified_code': ...}]"
},
"keep_recent": {
"type": "integer",
"default": 20
},
"max_new_skills": {
"type": "integer",
"default": 4
}
}
}Execute pure Python math/algorithmic verification inside Oxyo's isolated sandbox (with optional AST-driven adversarial boundary fuzzing & invariant checks). Assign the final output to RESULT = ...
| Argument | Type | Description |
|---|---|---|
| code* | string | Python snippet setting RESULT = ... |
| fuzz | boolean | Enable adversarial boundary/edge-case property fuzzing & determinism check |
| timeout_s | number | |
| entrypoint | string | Optional top-level function name to fuzz directly |
| invariants | array | Optional boolean Python expressions (e.g. 'RESULT > 0') that must hold |
{
"type": "object",
"required": [
"code"
],
"properties": {
"code": {
"type": "string",
"description": "Python snippet setting RESULT = ..."
},
"fuzz": {
"type": "boolean",
"default": false,
"description": "Enable adversarial boundary/edge-case property fuzzing & determinism check"
},
"timeout_s": {
"type": "number",
"default": 4
},
"entrypoint": {
"type": "string",
"description": "Optional top-level function name to fuzz directly"
},
"invariants": {
"type": "array",
"items": {
"type": "string"
},
"description": "Optional boolean Python expressions (e.g. 'RESULT > 0') that must hold"
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 11, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.
| chain_sequentially | boolean |
| acceptance_criteria | string |
| prospective_trigger | string |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 5
},
"scope": {
"type": "string"
},
"title": {
"type": "string"
},
"action": {
"enum": [
"create",
"decompose",
"next",
"update",
"tree"
],
"type": "string",
"default": "next"
},
"status": {
"enum": [
"pending",
"in_progress",
"completed",
"failed",
"blocked"
],
"type": "string"
},
"goal_id": {
"type": "string"
},
"evidence": {
"type": "string",
"default": ""
},
"priority": {
"type": "number",
"default": 0.75
},
"subgoals": {
"type": "array",
"items": {
"type": "object"
}
},
"parent_id": {
"type": "string"
},
"depends_on": {
"type": "array",
"items": {
"type": "string"
}
},
"failure_cause": {
"type": "string",
"default": ""
},
"chain_sequentially": {
"type": "boolean",
"default": false
},
"acceptance_criteria": {
"type": "string",
"default": ""
},
"prospective_trigger": {
"type": "string",
"default": ""
}
}
}