You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
-**deps:** bump fast-xml-parser from 5.4.1 to 5.4.2 ([91a1ff8](https://github.com/salesforcecli/plugin-agent/commit/91a1ff8124ed56b86c421077067d0e5f4a6ccee1))
-**deps:** bump @salesforce/core from 8.26.2 to 8.26.3 ([033b12c](https://github.com/salesforcecli/plugin-agent/commit/033b12cc8fab8a94091f50b524cc02d70a165861))
-**deps:** bump @salesforce/agents from 0.23.4 to 0.24.1 ([229e94c](https://github.com/salesforcecli/plugin-agent/commit/229e94cb85adb6bd910f4175a3c19a09c5610ce2))
- initial updates, use Org/Connection, make beta/hidden ([2a0f283](https://github.com/salesforcecli/plugin-agent/commit/2a0f28396cd31ae9c8069a303501d77d5eac6bac))
24
+
25
+
### Features
26
+
27
+
- add `sf agent test run-eval` for Einstein Eval Labs API testing ([1699d51](https://github.com/salesforcecli/plugin-agent/commit/1699d5188a3e4b165c564680e65e5dc33b20bb6e))
28
+
- add YAML test spec support for `sf agent test run-eval` ([e942e00](https://github.com/salesforcecli/plugin-agent/commit/e942e006e7892d34ae93e6f84a6c6673b8be0c9b))
29
+
- normalize MCP shorthand format in evalNormalizer ([01bc87b](https://github.com/salesforcecli/plugin-agent/commit/01bc87be643b4901ba65d79b205a1977f750c81a))
Execute rich evaluation tests against an Agentforce agent using the Einstein Evaluation API. Supports both YAML test specs (same format as `sf agent generate test-spec`) and JSON payloads.
8
+
9
+
When you provide a YAML test spec, the command automatically translates test cases into Evaluation API calls and infers the agent name from the spec's `subjectName` field. This means you can use the same test spec with both `sf agent test run` and `sf agent test run-eval`.
10
+
11
+
When you provide a JSON payload, it's sent directly to the API with optional normalization. The normalizer auto-corrects common field name mistakes, converts shorthand references to JSONPath, and injects defaults. Use `--no-normalize` to disable this auto-normalization.
12
+
13
+
Supports 8+ evaluator types, including topic routing assertions, action invocation checks, string/numeric assertions, semantic similarity scoring, and LLM-based quality ratings.
14
+
15
+
# flags.spec.summary
16
+
17
+
Path to test spec file (YAML or JSON). Use `-` for stdin.
18
+
19
+
# flags.api-name.summary
20
+
21
+
Agent DeveloperName (also called API name) to resolve agent_id and agent_version_id. Auto-inferred from the YAML spec's subjectName.
22
+
23
+
# flags.wait.summary
24
+
25
+
Number of minutes to wait for results.
26
+
27
+
# flags.result-format.summary
28
+
29
+
Format of the agent test results.
30
+
31
+
# flags.batch-size.summary
32
+
33
+
Number of tests per API request (max 5).
34
+
35
+
# flags.no-normalize.summary
36
+
37
+
Disable auto-normalization of field names and shorthand references.
38
+
39
+
# examples
40
+
41
+
- Run tests using a YAML test spec on the org with alias "my-org":
0 commit comments