CodeSpecWeave builds coding tasks with interdependent specifications
The pipeline produces 4,670 tasks across 2,865 repositories and a 100-task benchmark with 1,044 specifications.
Stories per day · 30 days
Relatedcoding-agents3agent-smith1benchmark1benchmarks1developer-tools1llm-agents1program-synthesis1pull-requests1reinforcement-learning1software-engineering1
3 stories, newest first · one card per story
Stories tagged #specifications
Skeleton Coding uses program structure to steer coding agents
Across four models and three benchmarks, the best skeletons cut mean test error 48.4% versus the best behavioral specs.
Spec-first plugin makes coding agents build from reviewed specs
The plugin gates commits on project checks and can build independent specs in parallel worktrees before one pull request.
1