Validate agent compatibility
POST /api/agents/architecture/validate/compatibility
Check whether two agents can safely run at the same time.
Inspects conflict declarations in both directions — if either agent lists the other in its conflicts_with set, they are incompatible, and conflicts_detected explains every side of the disagreement. Unknown agent IDs have no declared conflicts, so two unknown IDs are reported as compatible (matching the lenient behavior of the dependency/conflict lookups).
Args: request: CompatibilityRequest with agent_id_1 and agent_id_2.
Returns: CompatibilityResponse with are_compatible, any detected conflicts, and a message.
Request Body required
Section titled “Request Body required ”Request body for a pairwise agent compatibility check.
Unknown agent IDs are handled leniently — an ID not registered in the architecture has no declared conflicts, so a pair of unknown IDs will be reported as compatible. Callers that need strict existence checks should validate IDs against the architecture overview endpoint first. Empty strings are rejected at parse time via min_length=1.
object
First agent identifier
Second agent identifier
Responses
Section titled “ Responses ”Successful Response
Result of checking whether two agents can run simultaneously.
Inspects conflict declarations bidirectionally. conflicts_detected lists every side of the disagreement when are_compatible is False — both “a conflicts with b” and “b conflicts with a” appear if both are declared. Unknown IDs have no declared conflicts and therefore return are_compatible=True.
object
First agent identifier
Second agent identifier
Compatibility status
Specific conflicts found
Compatibility result message
Validation Error