Product #11 — Now Available

CAIN Skills

Governed capability layer for autonomous agents. Discover. Authorize. Execute. Evaluate. Improve.

Step 1

DRAFT

→

Step 2

VALIDATE

→

Step 3

SECURITY

→

Step 4

TEST

→

Step 5

APPROVE

→

Step 6

ACTIVE

System Statistics

Total Skills

Active

Pending

Approved

Suspended

Create New Skill

Author a new skill with full governance lifecycle.

Skill Registry

Browse and manage registered skills.

Execute Skill

Execute an active skill through CAIN governance.

Skill Governance Workflow

DISCOVER → IDENTIFY

Agent discovers skill in registry. Skill is identified and indexed by type, namespace, and capabilities.

PROVENANCE → VALIDATE → SECURITY SCREEN

Skill provenance is verified. Validation checks inputs/outputs schemas. Security scanner detects threats.

POLICY → TEST

Policy constraints are evaluated. Test suite validates skill behavior.

APPROVAL → ACTIVATE

Authorizer approves skill. Skill transitions to ACTIVE state.

EXECUTE → OBSERVE → EVALUATE

Skill is executed through CAIN enforcement pipeline. Execution is traced and evaluated.

UPDATE / REVOKE

Behavioral drift detection may trigger DEGRADED/SUSPENDED. Updates create new versions. Authorizer may revoke.

API Reference

GET /fabric/skills - List all skills
GET /fabric/skills/stats - Get statistics
POST /fabric/skills - Create skill
GET /fabric/skills/{id} - Get skill details
PATCH /fabric/skills/{id}/state - Update state
POST /fabric/skills/{id}/validate - Validate
POST /fabric/skills/{id}/security-scan - Security scan
POST /fabric/skills/{id}/test - Test
POST /fabric/skills/{id}/approve - Approve
POST /fabric/skills/{id}/activate - Activate
POST /fabric/skills/{id}/execute - Execute
POST /fabric/skills/{id}/suspend - Suspend
POST /fabric/skills/{id}/revoke - Revoke
GET /fabric/skills/{id}/versions - Get versions
GET /fabric/skills/{id}/provenance - Get provenance
GET /fabric/skills/{id}/executions - Get executions