agentv 2.5.5 → 2.5.7

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (8) hide show
  1. package/README.md +5 -11
  2. package/dist/{chunk-LTEYARQG.js → chunk-KM7W27VC.js} +1050 -354
  3. package/dist/chunk-KM7W27VC.js.map +1 -0
  4. package/dist/cli.js +1 -1
  5. package/dist/index.js +1 -1
  6. package/dist/templates/.claude/skills/agentv-eval-builder/SKILL.md +2 -2
  7. package/package.json +1 -1
  8. package/dist/chunk-LTEYARQG.js.map +0 -1
package/README.md CHANGED
@@ -31,13 +31,9 @@ evalcases:
31
31
  - id: addition
32
32
  expected_outcome: Correctly calculates 15 + 27 = 42
33
33
 
34
- input_messages:
35
- - role: user
36
- content: What is 15 + 27?
34
+ input: What is 15 + 27?
37
35
 
38
- expected_messages:
39
- - role: assistant
40
- content: "42"
36
+ expected_output: "42"
41
37
 
42
38
  execution:
43
39
  evaluators:
@@ -108,8 +104,8 @@ See [AGENTS.md](AGENTS.md) for development guidelines and design principles.
108
104
  For large-scale evaluations, AgentV supports JSONL (JSON Lines) format as an alternative to YAML:
109
105
 
110
106
  ```jsonl
111
- {"id": "test-1", "expected_outcome": "Calculates correctly", "input_messages": [{"role": "user", "content": "What is 2+2?"}]}
112
- {"id": "test-2", "expected_outcome": "Provides explanation", "input_messages": [{"role": "user", "content": "Explain variables"}]}
107
+ {"id": "test-1", "expected_outcome": "Calculates correctly", "input": "What is 2+2?"}
108
+ {"id": "test-2", "expected_outcome": "Provides explanation", "input": "Explain variables"}
113
109
  ```
114
110
 
115
111
  Optional sidecar YAML metadata file (`dataset.yaml` alongside `dataset.jsonl`):
@@ -264,9 +260,7 @@ evalcases:
264
260
  - id: quicksort-explain
265
261
  expected_outcome: Explain how quicksort works
266
262
 
267
- input_messages:
268
- - role: user
269
- content: Explain quicksort algorithm
263
+ input: Explain quicksort algorithm
270
264
 
271
265
  rubrics:
272
266
  - Mentions divide-and-conquer approach