curl --request POST \
--url https://api.mnemom.ai/v1/policies/evaluate \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"policy": {
"meta": {
"schema_version": "<string>",
"name": "<string>",
"description": "<string>",
"scope": "agent"
},
"capability_mappings": {},
"forbidden": [
{
"pattern": "<string>",
"reason": "<string>",
"severity": "high"
}
],
"escalation_triggers": [
{
"condition": "<string>",
"action": "notify",
"reason": "<string>"
}
],
"defaults": {
"unmapped_tool_action": "warn",
"unmapped_severity": "medium",
"fail_open": false,
"enforcement_mode": "observe",
"grace_period_hours": 0
}
},
"tools": [
"<string>"
],
"agent_id": "<string>"
}
'{
"verdict": "pass",
"violations": [
{
"type": "forbidden",
"tool": "<string>",
"capability": "<string>",
"rule": "<string>",
"reason": "<string>",
"severity": "low"
}
],
"warnings": [
{
"tool": "<string>",
"message": "<string>"
}
],
"card_gaps": [
{
"capability": "<string>",
"missing_card_field": "<string>",
"suggestion": "<string>"
}
],
"coverage": 0.5
}Evaluate a CLPI policy against a list of tools to determine which are allowed, forbidden, or unmapped. Returns a verdict with detailed violations, warnings, card coverage gaps, and overall coverage percentage. Useful for dry-run validation before deploying a policy.
curl --request POST \
--url https://api.mnemom.ai/v1/policies/evaluate \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"policy": {
"meta": {
"schema_version": "<string>",
"name": "<string>",
"description": "<string>",
"scope": "agent"
},
"capability_mappings": {},
"forbidden": [
{
"pattern": "<string>",
"reason": "<string>",
"severity": "high"
}
],
"escalation_triggers": [
{
"condition": "<string>",
"action": "notify",
"reason": "<string>"
}
],
"defaults": {
"unmapped_tool_action": "warn",
"unmapped_severity": "medium",
"fail_open": false,
"enforcement_mode": "observe",
"grace_period_hours": 0
}
},
"tools": [
"<string>"
],
"agent_id": "<string>"
}
'{
"verdict": "pass",
"violations": [
{
"type": "forbidden",
"tool": "<string>",
"capability": "<string>",
"rule": "<string>",
"reason": "<string>",
"severity": "low"
}
],
"warnings": [
{
"tool": "<string>",
"message": "<string>"
}
],
"card_gaps": [
{
"capability": "<string>",
"missing_card_field": "<string>",
"suggestion": "<string>"
}
],
"coverage": 0.5
}Supabase JWT token in Authorization: Bearer header
A complete CLPI policy document defining capability mappings, forbidden tools, escalation triggers, and enforcement defaults.
Show child attributes
List of tool identifiers to evaluate against the policy
Optional agent ID to include agent-specific context in evaluation
Policy evaluation result
Result of evaluating a CLPI policy against a set of tools or traces.
Overall evaluation verdict
pass, warn, fail List of policy violations detected
Show child attributes
Non-blocking warnings about potential issues
Show child attributes
Gaps between the policy and the alignment card
Show child attributes
Fraction of tools covered by the policy (0.0 to 1.0)
0 <= x <= 1