Ventures & Missions

Ventures are the top-level organizational containers in Kaneru. Each venture groups missions, projects, agents, directives, and cost budgets into a single manageable unit. Every state change is committed to the DAG for full auditability.

Creating a Venture

CLI

bash
mayros kaneru venture create \
  --name "Security Audit" \
  --prefix SEC \
  --directive "Ship secure code" \
  --fuel-limit 10000

Flag	Description	Required
`--name`	Venture display name	Yes
`--prefix`	Short prefix for mission IDs (e.g. `SEC-1`, `SEC-2`)	Yes
`--directive`	Top-level strategic directive	No
`--fuel-limit`	Maximum fuel budget in integer cents	No

MCP

Use the kaneru_venture_create tool from any MCP client (Claude Desktop, Claude Code, Cursor):

"Create a venture called Security Audit with prefix SEC and a fuel limit of 10000"
→ calls kaneru_venture_create

Portal Wizard

Open mayros dashboard and navigate to the Ventures tab. Click New Venture to launch the 4-step setup wizard:

Name & prefix — choose a name and a short identifier
Directives — define strategic goals as a tree
Chain of command — assign agents and set escalation
Fuel budget — set spending limits

Listing and Inspecting Ventures

bash
# List all ventures
mayros kaneru venture list

# Get detailed status for a venture
mayros kaneru venture status --venture <id>

The status command shows mission counts by state, fuel usage, active agents, and directive progress.

Mission Lifecycle

Missions are the work units agents execute. Each mission belongs to a venture and follows a strict state machine:

mermaid
stateDiagram-v2
    [*] --> queued
    queued --> ready
    ready --> active : agent claims
    active --> review
    active --> abandoned
    review --> complete
    complete --> [*]
    abandoned --> [*]

Creating Missions

bash
mayros kaneru mission create \
  --venture <venture-id> \
  --title "Audit auth module" \
  --priority critical \
  --description "Review OAuth flows for token leakage"

Valid priorities: critical, high, medium, low.

Claiming and Transitioning

Agents claim missions atomically with optimistic concurrency. Only the claiming run can transition the mission forward.

bash
# Claim a mission for an agent run
mayros kaneru mission claim --mission <id> --agent scanner --run run-001

# Transition to a new status
mayros kaneru mission transition --mission <id> --status review --run run-001

# Complete with learning update and knowledge transfer
mayros kaneru mission complete-with-learning \
  --mission <id> \
  --run run-001 \
  --domain typescript \
  --task-type security-scan \
  --score 0.95

The complete-with-learning command updates the agent's learning profile and transfers knowledge to the venture's shared namespace.

Listing Missions

bash
# All missions for a venture
mayros kaneru mission list --venture <id>

# Filter by status
mayros kaneru mission list --venture <id> --status active

Projects

Projects group related missions within a venture for organization and tracking.

bash
# Create a project
mayros kaneru project create \
  --venture <id> \
  --name "Q1 Security Sprint" \
  --description "All security-related missions for Q1"

# List projects
mayros kaneru project list --venture <id>

# Get project status (mission breakdown)
mayros kaneru project status --project <id>

Comments

Comments provide threaded discussion on missions, visible to all agents and human operators.

bash
# Add a comment to a mission
mayros kaneru comment add --mission <id> --author scanner --body "Found 3 issues in auth.ts"

# List comments for a mission
mayros kaneru comment list --mission <id>

Comments are stored as Cortex triples and included in the DAG audit trail.

Chain of Command

The chain of command defines agent hierarchy and escalation paths within a venture. When an agent encounters a problem beyond its scope, it escalates to the next agent in the chain.

CTO (strategic decisions)
├── Lead Engineer (architecture)
│   ├── Scanner (security audits)
│   └── Builder (implementation)
└── Ops Lead (deployment)
    └── Monitor (observability)

Set up the chain when creating a venture via the portal wizard, or manage it programmatically through the venture layer API.

Directives

Directives form a tree of goals from strategic to tactical:

Strategic: "Ship secure code"
├── Objective: "Zero critical CVEs in production"
│   ├── Task: "Audit all OAuth flows"
│   └── Task: "Add rate limiting to API"
└── Objective: "OWASP Top 10 compliance"
    ├── Task: "SQL injection scan"
    └── Task: "XSS prevention review"

Directives guide agent behavior — when routing missions, the system considers which agents align with which directives. Create directives during venture setup or add them later through the API.

Next Steps

Learning & Routing — how agents build expertise
DAG Audit Trail — immutable history of every action
Kaneru CLI Reference — all 18 subcommand groups
MCP Tools Reference — all 24 tools