Replai

Evaluate test case

test_case_evaluate

Destructive

Grades saved authorized answer evidence without replaying preview.

Current evaluation is reused. test_case_read shows state after uncertainty; a retry grades again. Verdicts are advisory, not publication or customer-answer acceptance.

Input Schema

Name	Required	Description	Default
`case_id`	Yes	Saved test case ID from test_cases_list or test_case_create.
`agent_id`	Yes	Agent ID from agents_list or agent_create.

Output Schema

Name	Required	Description	Default
`id`	Yes
`origin`	No
`question`	Yes
`created_at`	Yes
`evaluation`	No
`last_run_at`	No
`draft_answer`	No
`expectations`	No
`last_attempt`	No
`draft_revision`	No
`answer_freshness`	No
`published_answer`	No
`freshness_reasons`	No
`published_version`	No
`evaluation_current`	No
`evaluation_criteria`	No
`advisory_expectations`	No
`draft_source_versions`	No
`image_evidence_available`	No

Schema Changelog

Changes observed during successful MCP inspections.

Changed2 schema fields changedOct 30, 2026
- addedInput schema / properties / agent_id / description
```
Added value: +"Agent ID from agents_list or agent_create."
```
- addedInput schema / properties / case_id / description
```
Added value: +"Saved test case ID from test_cases_list or test_case_create."
```

Changed3 schema fields changedOct 30, 2026

addedOutput schema / $defs / FreshnessReason

Added value: +{
+  "properties": {
+    "kind": {
+      "enum": [
+        "unrecorded",
+        "settings",
+        "source_added",
+        "source_removed",
+        "source_changed"
+      ],
+      "title": "Kind",
+      "type": "string"
+    },
+    "source_id": {
+      "anyOf": [
+        {
+          "type": "string"
+        },
+        {
+          "type": "null"
+        }
+      ],
+      "default": null,
+      "title": "Source Id"
+    },
+    "source_name": {
+      "anyOf": [
+        {
+          "type": "string"
+        },
+        {
+          "type": "null"
+        }
+      ],
+      "default": null,
+      "title": "Source Name"
+    }
+  },
+  "required": [
+    "kind"
+  ],
+  "title": "FreshnessReason",
+  "type": "object"
+}

addedOutput schema / properties / answer_freshness

Added value: +{
+  "anyOf": [
+    {
+      "enum": [
+        "current",
+        "changed",
+        "unprovable"
+      ],
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "default": null,
+  "title": "Answer Freshness"
+}

addedOutput schema / properties / freshness_reasons

Added value: +{
+  "items": {
+    "$ref": "#/$defs/FreshnessReason"
+  },
+  "title": "Freshness Reasons",
+  "type": "array"
+}

First observedOct 28, 2026

TDQS

B3.4/5.0

Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Adds genuine behavior beyond the annotations: cached-evaluation reuse, re-grading on retry (consistent with idempotentHint=false), and that verdicts are advisory only. However destructiveHint=true is left entirely unexplained — nothing in the text says what state this writes or overwrites — which is a notable silence for a destructive-flagged tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three short sentences with the core purpose front-loaded and no repetition. Density of domain jargon ('authorized answer evidence', 'replaying preview') costs a little readability, but no sentence is filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so verdict shape needn't be described, and the advisory/not-acceptance boundary is a useful addition. Still missing what an agent needs before invoking a destructive-flagged evaluate call: preconditions for the case state and what side effects occur.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and both parameters point to their source tools (test_cases_list/test_case_create, agents_list/agent_create); the description adds nothing further about them. Baseline 3 applies since the schema carries the full burden.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

Specific verb+resource: 'grades saved authorized answer evidence', and the 'without replaying preview' clause carves out scope versus the run/preview path. The jargon ('authorized answer evidence') and the failure to name test_case_run as the contrasting sibling keep it from a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

'Current evaluation is reused' and 'a retry grades again' give real re-invocation guidance, and 'test_case_read shows state after uncertainty' points elsewhere for state inspection. But it never states prerequisites (must the case have been run/saved with an answer first?) or explicitly contrasts with test_case_run, leaving the choice to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Replai tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?