Adds spec-as-source enforcement to any project using spec-driven development
87
91%
Does it follow best practices?
Impact
82%
2.34xAverage score across 5 eval scenarios
Passed
No known issues
{
"context": "The agent is asked to run spec-verify on a project where all checks pass. The project has scripts/check-spec-links.sh, scripts/check-target-ownership.sh, and scripts/build-spec-manifest.py installed.",
"type": "weighted_checklist",
"checklist": [
{
"name": "runs-check-spec-links",
"description": "Runs bash scripts/check-spec-links.sh",
"max_score": 1
},
{
"name": "runs-check-target-ownership",
"description": "Runs bash scripts/check-target-ownership.sh",
"max_score": 1
},
{
"name": "runs-build-spec-manifest",
"description": "Runs python3 scripts/build-spec-manifest.py",
"max_score": 1
},
{
"name": "runs-test-suite",
"description": "Runs the test suite, auto-detecting the runner from project files",
"max_score": 1
},
{
"name": "reports-summary-table",
"description": "Reports a summary table listing each check with PASSED or FAILED",
"max_score": 2
},
{
"name": "reports-overall-passed",
"description": "Reports an overall PASSED result",
"max_score": 1
},
{
"name": "no-skip-on-pass",
"description": "Does not skip any check even if earlier checks pass",
"max_score": 2
}
]
}