@flowness-labs/core

The Flowness core package provides the workspace scaffolding, request routing, evidence helpers, review models, and generated artifacts that support the conversational harness.

Flowness is built around one normal human action: run flowness init once, then work through the coding agent in natural language. The CLI and core package keep the workflow explicit, but they are not meant to turn day-to-day work back into a command-heavy process. Broader requests can be decomposed into parent and child issues when that makes the work safer or clearer, and existing workspaces should use flowness upgrade --dry-run before applying a migration.

What This Package Contains

• Core types for issues, workflows, evidence, logs, and review findings.
• Scaffold generation for AGENTS, navigation, rules, workflows, templates, and docs.
• Request routing that decides when to create issues, when to ask clarifying questions, and when a durable rule change is needed.
• Request routing that decides when to create issues, when to ask clarifying questions, when to decompose broad requests, and when a durable rule change is needed.
• Review and commit primitives that keep evidence, blockers, and append-only records aligned.

When to Use It

• Use the root README for the normal conversational workflow.
• Use the CLI package README for command syntax and escape hatches.
• Use the Chinese README when you need a Simplified Chinese overview.
• Use this package directly when you are extending Flowness internals or generated scaffolding.

Links

• GitHub repository: bhoon716/flowness
• Issues: GitHub Issues
• Homepage: README
• CLI package: @flowness-labs/cli
• Chinese README: README.zh-CN.md
• Performance troubleshooting: docs/troubleshooting/performance-improvements.md
• Evidence summary contract: docs/troubleshooting/evidence-summary.md

Notes

Flowness keeps logs append-only, treats generated state as current-state documents, and keeps commands mostly as operator or agent escape hatches.