Skip to content

Commit 43cc12e

Browse files
committed
add tags for all single-case eval files and test workaround in no-skill pipeline
1 parent bcb4acf commit 43cc12e

31 files changed

Lines changed: 2779 additions & 1420 deletions

.github/skills/azure-typespec-author/evaluate/evals/001001.eval.yaml

Lines changed: 78 additions & 40 deletions
Original file line numberDiff line numberDiff line change
@@ -16,61 +16,99 @@ config:
1616
# Test cases
1717
stimuli:
1818
- name: 001001-version-spread-property
19-
prompt: I added `...Azure.ResourceManager.ManagedServiceIdentityProperty;` which updates all my existing API versions and introduces a breaking change. I want to introduce the properties of the spread model '...ManagedServiceIdentityProperty' in API version 2025-05-04-preview only.
20-
environment:
19+
tags:
20+
- 'main'
21+
prompt: 'I added `...Azure.ResourceManager.ManagedServiceIdentityProperty;` which updates all my existing API versions and introduces a breaking change. I want to introduce the properties of the spread model ''...ManagedServiceIdentityProperty'' in API version 2025-05-04-preview only.'
22+
environment: &shared_environment
2123
files:
22-
- src: ../fixtures/001-share-version-new-feature/employee.tsp
23-
dest: employee.tsp
24-
- src: ../fixtures/001-share-version-new-feature/main.tsp
25-
dest: main.tsp
26-
- src: ../fixtures/Microsoft.Widget/Widget/package-lock.json
27-
dest: package-lock.json
28-
- src: ../fixtures/Microsoft.Widget/Widget/package.json
29-
dest: package.json
30-
- src: ../fixtures/001-share-version-new-feature/readme.md
31-
dest: readme.md
32-
- src: ../fixtures/001-share-version-new-feature/shared.tsp
33-
dest: shared.tsp
34-
- src: ../fixtures/001-share-version-new-feature/tspconfig.yaml
35-
dest: tspconfig.yaml
24+
- src: ../fixtures/001-share-version-new-feature/employee.tsp
25+
dest: employee.tsp
26+
- src: ../fixtures/001-share-version-new-feature/main.tsp
27+
dest: main.tsp
28+
- src: ../fixtures/Microsoft.Widget/Widget/package-lock.json
29+
dest: package-lock.json
30+
- src: ../fixtures/Microsoft.Widget/Widget/package.json
31+
dest: package.json
32+
- src: ../fixtures/001-share-version-new-feature/readme.md
33+
dest: readme.md
34+
- src: ../fixtures/001-share-version-new-feature/shared.tsp
35+
dest: shared.tsp
36+
- src: ../fixtures/001-share-version-new-feature/tspconfig.yaml
37+
dest: tspconfig.yaml
3638
commands:
37-
- node -e "const fs=require('fs');const t=process.env.FIXTURE_NODE_MODULES;if(t&&fs.existsSync(t)&&!fs.existsSync('node_modules')){fs.symlinkSync(t,'node_modules','junction')}"
39+
- 'node -e "const fs=require(''fs'');const t=process.env.FIXTURE_NODE_MODULES;if(t&&fs.existsSync(t)&&!fs.existsSync(''node_modules'')){fs.symlinkSync(t,''node_modules'',''junction'')}"'
3840
graders:
39-
- type: tool-calls
41+
- type: 'tool-calls'
4042
config:
4143
required:
42-
- edit
43-
- azure-sdk-mcp-azsdk_typespec_generate_authoring_plan
44-
- azure-sdk-mcp-azsdk_run_typespec_validation
45-
- type: skill-invocation
44+
- 'edit'
45+
- 'azure-sdk-mcp-azsdk_typespec_generate_authoring_plan'
46+
- 'azure-sdk-mcp-azsdk_run_typespec_validation'
47+
- type: 'skill-invocation'
4648
config:
4749
required:
48-
- azure-typespec-author
49-
- type: file-matches
50+
- 'azure-typespec-author'
51+
- type: 'file-matches'
5052
config:
51-
path: main.tsp
52-
pattern: 2025-05-04-preview
53-
- type: file-matches
53+
pattern: '2025-05-04-preview'
54+
path: 'main.tsp'
55+
- type: 'file-matches'
5456
config:
55-
path: employee.tsp
56-
pattern: "\\.\\.\\.ManagedServiceIdentityProperty;"
57-
- type: file-matches
57+
pattern: '\.\.\.ManagedServiceIdentityProperty;'
58+
path: 'employee.tsp'
59+
- type: 'file-matches'
5860
config:
59-
path: employee.tsp
60-
pattern: "@@added\\(Employee\\.identity, (Microsoft\\.Widget\\.)?Versions\\.(v2025_05_04_preview|`2025-05-04-preview`)\\);"
61-
- type: prompt
61+
pattern: '@@added\(Employee\.identity, (Microsoft\.Widget\.)?Versions\.(v2025_05_04_preview|`2025-05-04-preview`)\);'
62+
path: 'employee.tsp'
63+
- type: 'prompt'
6264
config:
63-
prompt: Verify changes are scoped to this task only, with no unrelated edits beyond introducing preview-only identity via versioning.
64-
model: claude-opus-4.6
65-
scoring: scale_1_5
66-
threshold: 1.0
65+
prompt: 'Verify changes are scoped to this task only, with no unrelated edits beyond introducing preview-only identity via versioning.'
66+
model: 'claude-opus-4.6'
67+
threshold: 1
68+
scoring: 'scale_1_5'
69+
constraints:
70+
max_turns: 5
71+
max_tokens: 50000
72+
- name: 001001-version-spread-property-trigger
73+
tags:
74+
- 'trigger'
75+
prompt: 'I added `...Azure.ResourceManager.ManagedServiceIdentityProperty;` which updates all my existing API versions and introduces a breaking change. I want to introduce the properties of the spread model ''...ManagedServiceIdentityProperty'' in API version 2025-05-04-preview only.'
76+
environment: *shared_environment
77+
graders:
78+
- type: 'tool-calls'
79+
config:
80+
required:
81+
- 'edit'
82+
- 'azure-sdk-mcp-azsdk_run_typespec_validation'
83+
- type: 'skill-invocation'
84+
config:
85+
required:
86+
- 'azure-typespec-author'
87+
constraints:
88+
max_turns: 5
89+
max_tokens: 50000
90+
- name: 001001-version-spread-property-forced
91+
tags:
92+
- 'forced'
93+
prompt: '@azure-typespec-author I added `...Azure.ResourceManager.ManagedServiceIdentityProperty;` which updates all my existing API versions and introduces a breaking change. I want to introduce the properties of the spread model ''...ManagedServiceIdentityProperty'' in API version 2025-05-04-preview only.'
94+
environment: *shared_environment
95+
graders:
96+
- type: 'tool-calls'
97+
config:
98+
required:
99+
- 'edit'
100+
- 'azure-sdk-mcp-azsdk_run_typespec_validation'
101+
- type: 'skill-invocation'
102+
config:
103+
required:
104+
- 'azure-typespec-author'
67105
constraints:
68106
max_turns: 5
69107
max_tokens: 50000
70108
scoring:
71109
weights:
72-
file-matches: 3
110+
prompt: 1
73111
tool-calls: 1
74112
skill-invocation: 1
75-
prompt: 1
76-
threshold: 1.0
113+
file-matches: 3
114+
threshold: 1

.github/skills/azure-typespec-author/evaluate/evals/001002.eval.yaml

Lines changed: 82 additions & 39 deletions
Original file line numberDiff line numberDiff line change
@@ -16,66 +16,109 @@ config:
1616
# Test cases
1717
stimuli:
1818
- name: 001002-version-default-value
19+
tags:
20+
- 'main'
1921
prompt: |
2022
add a default value `21` for property age in model EmployeeProperties only for the new api version 2025-11-01
21-
constraints:
22-
max_turns: 6
23-
max_tokens: 100000
24-
environment:
23+
24+
environment: &shared_environment
2525
files:
26-
- src: ../fixtures/001002-version-default-value/employee.tsp
27-
dest: employee.tsp
28-
- src: ../fixtures/001002-version-default-value/main.tsp
29-
dest: main.tsp
30-
- src: ../fixtures/Microsoft.Widget/Widget/package-lock.json
31-
dest: package-lock.json
32-
- src: ../fixtures/Microsoft.Widget/Widget/package.json
33-
dest: package.json
34-
- src: ../fixtures/001002-version-default-value/readme.md
35-
dest: readme.md
36-
- src: ../fixtures/001002-version-default-value/shared.tsp
37-
dest: shared.tsp
38-
- src: ../fixtures/001002-version-default-value/tspconfig.yaml
39-
dest: tspconfig.yaml
26+
- src: ../fixtures/001002-version-default-value/employee.tsp
27+
dest: employee.tsp
28+
- src: ../fixtures/001002-version-default-value/main.tsp
29+
dest: main.tsp
30+
- src: ../fixtures/Microsoft.Widget/Widget/package-lock.json
31+
dest: package-lock.json
32+
- src: ../fixtures/Microsoft.Widget/Widget/package.json
33+
dest: package.json
34+
- src: ../fixtures/001002-version-default-value/readme.md
35+
dest: readme.md
36+
- src: ../fixtures/001002-version-default-value/shared.tsp
37+
dest: shared.tsp
38+
- src: ../fixtures/001002-version-default-value/tspconfig.yaml
39+
dest: tspconfig.yaml
4040
commands:
41-
- node -e "const fs=require('fs');const t=process.env.FIXTURE_NODE_MODULES;if(t&&fs.existsSync(t)&&!fs.existsSync('node_modules')){fs.symlinkSync(t,'node_modules','junction')}"
41+
- 'node -e "const fs=require(''fs'');const t=process.env.FIXTURE_NODE_MODULES;if(t&&fs.existsSync(t)&&!fs.existsSync(''node_modules'')){fs.symlinkSync(t,''node_modules'',''junction'')}"'
4242
graders:
43-
- type: tool-calls
43+
- type: 'tool-calls'
4444
config:
4545
required:
46-
- edit
47-
- azure-sdk-mcp-azsdk_typespec_generate_authoring_plan
48-
- azure-sdk-mcp-azsdk_run_typespec_validation
49-
- type: skill-invocation
46+
- 'edit'
47+
- 'azure-sdk-mcp-azsdk_typespec_generate_authoring_plan'
48+
- 'azure-sdk-mcp-azsdk_run_typespec_validation'
49+
- type: 'skill-invocation'
5050
config:
5151
required:
52-
- azure-typespec-author
53-
- type: file-matches
52+
- 'azure-typespec-author'
53+
- type: 'file-matches'
5454
config:
55-
path: employee.tsp
5655
pattern: '@removed\(Versions\.v2025_11_01\)'
57-
- type: file-matches
56+
path: 'employee.tsp'
57+
- type: 'file-matches'
5858
config:
59-
path: employee.tsp
6059
pattern: '@renamedFrom\(Versions\.v2025_11_01, "age"\)'
61-
- type: file-matches
60+
path: 'employee.tsp'
61+
- type: 'file-matches'
6262
config:
63-
path: employee.tsp
6463
pattern: '@added\(Versions\.v2025_11_01\)'
65-
- type: file-matches
64+
path: 'employee.tsp'
65+
- type: 'file-matches'
6666
config:
67-
path: employee.tsp
6867
pattern: 'age\?: int32 = 21;'
69-
- type: prompt
68+
path: 'employee.tsp'
69+
- type: 'prompt'
7070
config:
71-
scoring: scale_1_5
72-
threshold: 1.0
71+
threshold: 1
72+
model: 'claude-opus-4.6'
73+
scoring: 'scale_1_5'
7374
prompt: 'Verify change scope and quality only: add a default value of 21 for the `age` property in `EmployeeProperties` that only applies to the 2025-11-01 API version. The old age renamed and removed for the new version.'
74-
model: claude-opus-4.6
75+
constraints:
76+
max_turns: 6
77+
max_tokens: 100000
78+
- name: 001002-version-default-value-trigger
79+
tags:
80+
- 'trigger'
81+
prompt: |
82+
add a default value `21` for property age in model EmployeeProperties only for the new api version 2025-11-01
83+
84+
environment: *shared_environment
85+
graders:
86+
- type: 'tool-calls'
87+
config:
88+
required:
89+
- 'edit'
90+
- 'azure-sdk-mcp-azsdk_run_typespec_validation'
91+
- type: 'skill-invocation'
92+
config:
93+
required:
94+
- 'azure-typespec-author'
95+
constraints:
96+
max_turns: 5
97+
max_tokens: 50000
98+
- name: 001002-version-default-value-forced
99+
tags:
100+
- 'forced'
101+
prompt: |
102+
@azure-typespec-author add a default value `21` for property age in model EmployeeProperties only for the new api version 2025-11-01
103+
104+
environment: *shared_environment
105+
graders:
106+
- type: 'tool-calls'
107+
config:
108+
required:
109+
- 'edit'
110+
- 'azure-sdk-mcp-azsdk_run_typespec_validation'
111+
- type: 'skill-invocation'
112+
config:
113+
required:
114+
- 'azure-typespec-author'
115+
constraints:
116+
max_turns: 5
117+
max_tokens: 50000
75118
scoring:
76119
weights:
120+
prompt: 1
77121
tool-calls: 1
78122
skill-invocation: 1
79-
prompt: 1
80123
file-matches: 4
81-
threshold: 1.0
124+
threshold: 1

0 commit comments

Comments
 (0)