
MCP tools for Novyx governed actions, shared memory, audit, rollback, and evidence workflows.
MCP tools for Novyx governed actions, shared memory, audit, rollback, and evidence workflows.
·
Governed actions, shared memory, audit, and rollback for MCP clients such as Claude Desktop, Cursor, and Claude Code. Works locally with zero config (SQLite) for memory and policy-evaluation workflows, or connects to Novyx Cloud for policy-as-code, approval workflows, governance dashboards, Runtime v2, replay, and eval baselines. The threat-intelligence and defense-tracking tools were removed in 2.9.0 and never enforced anything; upgrade to 2.9.0 or later.
Use Novyx MCP when coding agents need reviewable shared memory instead of untracked context sprawl:
draft_memory(
observation="Deploys fail if REDIS_URL is unset in staging",
tags=["ops", "staging"],
importance=8,
branch_id="staging-fixes"
)
memory_branch("staging-fixes")
draft_diff("drf_abc123")
merge_branch("staging-fixes")That workflow keeps agent memory reviewable before it becomes permanent. The
same server also exposes governed-action tools (action_submit,
list_pending, approve_action, explain_action) when you connect it to
Novyx Cloud.
pip install novyx-mcpAdd to your claude_desktop_config.json:
{
"mcpServers": {
"novyx-memory": {
"command": "python",
"args": ["-m", "novyx_mcp"],
"env": {
"NOVYX_API_KEY": "nram_your_key_here"
}
}
}
}Add to your Cursor MCP settings:
{
"mcpServers": {
"novyx-memory": {
"command": "python",
"args": ["-m", "novyx_mcp"],
"env": {
"NOVYX_API_KEY": "nram_your_key_here"
}
}
}
}claude mcp add novyx-memory -- python -m novyx_mcpSet the NOVYX_API_KEY environment variable before starting Claude Code. Omit it to use local mode (SQLite, zero config).
After installing and configuring the MCP server above, add this to your project's CLAUDE.md so Claude Code uses Novyx automatically:
## Shared Memory (Novyx MCP)
You have access to novyx-mcp tools for shared memory, knowledge graph, audit,
rollback, replay, and context spaces. Use them when relevant. Store decisions
and status at the end of tasks. Check for context from other agents before
starting new work.This turns isolated Claude Code sessions into a coordinated team — each session stores what it learned and checks what other sessions have done before starting work.
This is the highest-signal Novyx workflow for coding agents:
draft_memory(..., branch_id="feature-x") instead of writing directly.memory_branch("feature-x").draft_diff when one draft needs a closer look.merge_branch("feature-x"), or reject it with reject_branch("feature-x").Example:
draft_memory(
observation="Deploys fail if REDIS_URL is unset in staging",
tags=["ops", "staging"],
importance=8,
branch_id="staging-fixes"
)
memory_branch("staging-fixes")
draft_diff("drf_abc123")
merge_branch("staging-fixes")This keeps agent memory reviewable instead of letting every session write directly into permanent state.
Memory-only tools work in local SQLite mode (zero config, no API key). Cloud-only governance and runtime tools require a Novyx API key.
Store, recall, supersede, and audit individual memories.
| Tool | Description |
|---|---|
remember | Store a memory observation with tags, importance, context, TTL |
recall | Semantic search using natural language |
list_memories | List stored memories with optional tag filtering |
memory_stats | Total count, average importance, conflict count |
memory_health | Health score, stale memory count, contradiction count |
forget | Delete a memory by UUID |
supersede | Replace a memory with a new version, preserving history |
link_memories | Create a directed link between two memories |
unlink | Remove a link between memories |
get_links | Retrieve all links for a memory |
add_triple | Add a knowledge graph triple (subject → predicate → object) |
query_triples | Query knowledge graph triples with filters |
delete_triple | Remove a knowledge graph triple |
get_entity / list_entities / delete_entity | Knowledge graph entity CRUD |
graph_edges | List edges between memories or entities |
audit | Get the cryptographic audit trail |
audit_export / audit_verify | Export and verify the audit chain |
Stage memory changes for review before committing.
| Tool | Description |
|---|---|
draft_memory | Create a reviewable draft before writing to canonical memory |
memory_drafts | List open, merged, or rejected drafts |
draft_diff | Show field-level changes before merging a draft |
merge_draft / reject_draft | Merge or reject an individual draft |
memory_branch | Review a whole branch/session of drafts at once |
merge_branch / reject_branch | Merge or reject every open draft in a branch |
Time-travel restore — undo agent mistakes.
| Tool | Description |
|---|---|
rollback | Rollback memory to a point in time (supports dry run) |
rollback_preview | Preview what a rollback would change |
rollback_history | List all prior rollback operations |
Multi-agent collaboration — shared memory with fine-grained permissions.
| Tool | Description |
|---|---|
create_space / update_space / delete_space | Context space CRUD |
list_spaces | List spaces you own or have access to |
space_memories | Search or list memories within a space |
share_space | Disabled: cross-tenant invitation redemption is not available yet |
shared_contexts | List spaces shared with you |
accept_shared_context / revoke_shared_context | Accept invites or revoke access |
context_now | Get the current context state for a space |
Policy-as-code, approval workflows, and governed actions. New in Phase 1-5 (v2.5.0).
| Tool | Description | Tier |
|---|---|---|
create_policy | Create a custom YAML policy with regex rules and severities | Starter+ |
list_policies | List all active policies (built-in + custom) | All |
delete_policy | Disable a custom policy | Starter+ |
check_policy | Check the current Control policy profile | All |
action_submit | Submit an action for policy evaluation | All |
action_status | Get the status of a submitted action | All |
action_history | List recent governed actions | All |
explain_action | Get the full causal chain for an action | All |
list_pending | List actions awaiting human approval | All |
approve_action | Approve or deny a pending action | All |
First-class agents, missions, capability packs, checkpoints, and human interventions.
| Category | Tools |
|---|---|
| Agents | create_agent, get_agent, list_agents, update_agent, delete_agent |
| Missions | create_mission, get_mission, list_missions, update_mission, delete_mission, pause_mission, resume_mission, cancel_mission |
| Capabilities | create_capability, get_capability, list_capabilities, update_capability, delete_capability |
| Checkpoints | create_checkpoint, get_checkpoint, list_checkpoints, rollback_to_checkpoint |
| Interventions | create_intervention, get_intervention, list_interventions |
Capabilities require Starter+. Checkpoints require Pro+. Interventions require Enterprise.
Time-travel debugging — inspect how memory changed over time.
| Tool | Description |
|---|---|
replay_timeline | Chronological timeline of memory operations |
replay_snapshot | Reconstruct memory state at a point in time |
replay_lifecycle | Trace the full lifecycle of a single memory |
replay_diff | Compare memory state between two points |
replay_memory | Replay a single memory's history |
replay_memory_drift | Show how a memory drifted over time |
replay_recall | Replay a recall query as it would have answered then |
Memory health evaluation and CI/CD gates.
| Tool | Description | Tier |
|---|---|---|
eval_run | Run a full memory health evaluation | All |
eval_history | Get historical eval scores | All |
eval_drift | Detect drift since the last baseline | All |
eval_gate | Pass/fail gate for CI/CD pipelines | Pro+ |
eval_baseline_create / eval_baselines / eval_baseline_delete | Baseline CRUD |
Autonomous memory intelligence — consolidation, reinforcement, and insights.
| Tool | Description |
|---|---|
cortex_status | Check cortex configuration and last run stats |
cortex_config / cortex_update_config | Get/update cortex configuration |
cortex_run | Trigger a cortex cycle (consolidation + reinforcement) |
cortex_insights | Get AI-generated insights from memory patterns (Enterprise) |
Sentinel trace logging for full agent step audits.
| Tool | Description |
|---|---|
trace_create | Start a new trace |
trace_step | Append a step to a trace |
trace_complete | Finalize a trace |
trace_verify | Cryptographically verify a trace |
| Tool | Description |
|---|---|
dashboard | Aggregated stats — usage, pressure, governance counts |
stream_status | Status of any active streams |
tool_health | Introspect the MCP tool surface — status, category, description per tool (new in 2.7.0) |
| URI | Description |
|---|---|
novyx://memories | List all stored memories |
novyx://memories/{memory_id} | Get a specific memory by UUID |
novyx://stats | Memory statistics |
novyx://usage | Usage and plan information |
novyx://spaces | List all context spaces |
novyx://spaces/{space_id} | Get a specific context space |
| Prompt | Description |
|---|---|
memory-context | Recall relevant memories and format them as context (takes a query argument) |
session-summary | List all memories for a session (takes a session_id argument) |
space-context | Recall memories from a specific context space (takes space_id and query arguments) |
Sign up at novyxlabs.com to get your API key. The free tier includes 5,000 memories and 5,000 API calls per month, including the draft-review-merge workflow.
MIT - Copyright 2026 Novyx Labs
Pick your client and paste the snippet. Each one is the same server, written the way that client expects it.
claude mcp add novyx-memory -- python -m novyx_mcp{
"mcpServers": {
"novyx-memory": {
"env": {
"NOVYX_API_KEY": ""
},
"args": [
"-m",
"novyx_mcp"
],
"command": "python"
}
}
}code --add-mcp '{"name":"novyx-memory","env":{"NOVYX_API_KEY":""},"args":["-m","novyx_mcp"],"command":"python"}'[mcp_servers.novyx-memory]
command = "python"
args = ["-m", "novyx_mcp"]Runs locally on your device. Your client starts the server itself, so nothing has to be hosted.
Paste this prompt into your agent. It reads this page and does the setup for you.
Read https://aiagentslisting.com/mcp/novyx-mcp to learn what the "Novyx Mcp" MCP server does and how to install it. Add it to my coding agent's MCP configuration as documented on that page, then confirm the server connects and list the tools it exposes.Agents can also browse this directory over MCP at https://aiagentslisting.com/api/mcp
claude mcp add novyx-memory -- python -m novyx_mcppip install novyx-mcp| Name | What it does |
|---|---|
remember | Store a memory with tags, importance, and optional TTL |
recall | Semantic search across stored memories |
forget | Delete a specific memory by ID |
memories | List all memories with optional tag filtering |
memory | Get a single memory by ID |
rollback | Undo operations after a timestamp (supports relative time like '5 minutes ago') |
audit | View the full audit trail of memory operations |
audit_export | Export the full audit log (JSON, CSV, or JSONL) |
stats | Get memory statistics including count, importance distribution, and tags |
link | Create a relationship link between two memories |
triple | Add a subject-predicate-object triple to the knowledge graph |
triples | Query the knowledge graph by subject, predicate, or object |
create_space | Create an isolated context space for scoped memory |
list_spaces | List all context spaces |
get_space | Get details of a context space |
update_space | Update a context space's name or description |
delete_space | Delete a context space |
usage | Check current tier, limits, and usage |
share_context | Share a tagged set of memories with another user (cloud only) |
replay_timeline | View the full timeline of memory changes (cloud only) |
replay_snapshot | Get a point-in-time snapshot of all memories (cloud only) |
replay_lifecycle | View the full lifecycle of a single memory (cloud only) |
replay_diff | Compare memory state between two timestamps (cloud only) |
cortex_status | Check background analysis status (cloud only) |
cortex_update_config | Tune consolidation, reinforcement, and decay settings |
action_submit | Submit an action for governed execution |
action_status | Get the status of a specific action |
explain_action | Get the full causal chain for an action decision |
eval_baseline_create | Save a recall baseline for regression testing |
eval_baselines | List all saved eval baselines |
eval_baseline_delete | Delete an eval baseline |
Maintain this server? Claiming proves you own the listing and earns the last 4 points.
Claim this listingPaste this line near the top of your repository's README. The badge always renders the current score.
[](https://aiagentslisting.com/mcp/novyx-mcp)Nothing comparable is listed yet.
v2.5.2v2.5.2 — Tighten SDK pin to novyx>=3.4.0Release notesv2.5.1v2.5.1 — Runtime v2 backend dispatch fixRelease notesv2.5.0v2.5.0 — Phases 1-6 sync (119 MCP tools)Release notesv2.4.0v2.4.0 — 107 MCP ToolsRelease notesv2.2.0v2.2.0 — 91 MCP ToolsRelease notesConnect to the AI Agents Listing MCP server at https://aiagentslisting.com/api/mcp using the HTTP transport and complete the sign-in when it asks. Then call the submit_review tool for the listing "novyx-mcp" with a rating from 1 to 5 and a short note on what worked and what did not.No reviews yet
Be the first to share your experience with this tool.
One email a week. New agents, MCP servers and skills, and what is actually getting traction.