openPrompting v1.0.0
openPrompting 1.0.0
The first stable release delivers a local, deterministic CLI for source-backed model, harness, and task guidance.
Highlights
- Configure one or more model/harness setups, named profiles, defaults, and task routes.
- Resolve guidance and generate structured prompt skeletons for five recurring task types.
- Diagnose local configuration, routing, instruction files, declared scripts, and guidance freshness without executing project code.
- Compare profiles, setups, models, and harnesses without inventing a universal ranking.
- Ship verified initial coverage for OpenAI/Codex, Anthropic/Claude Code, and Google/Gemini CLI.
- Validate knowledge, schemas, mixed-model routing, package contents, and packed-artifact execution in release gates.
Stable contracts
The executable, six command names, config schema v1, evidence vocabulary, and key data directories are frozen. See V1 contracts and migration notes.
Privacy and operation
Normal operation requires no account, hosted service, API key, model call, telemetry, or network connection. Setup writes only its project-local config; doctor is read-only.
Deferred beyond V1
Hosted documentation, native harness adapters, handoffs/workflows, prompt optimization, remote knowledge updates, broader harness coverage, and benchmarking remain in the post-V1 backlog.