robbond
39728667a7
- Add read_discovery_answers() to parse discovery-log.md rows into dicts with low_confidence and needs_followup boolean flags. - Add append_followup_answer() to append follow-up rows linked to original Q-A IDs. - Add 'ask-more' CLI command that detects Low-confidence and Follow-up-needed answers, prompts for additional details, and records them in the log. - Add 6 tests: read_discovery_answers round-trip, empty-log handling, low-confidence detection, follow-up flag detection, and append_followup.
rdb-discovery
A small CLI-first tool for repeatable software project discovery and delivery.
It creates markdown files that help humans and AI coding agents understand:
- what is being built
- why it exists
- what questions remain open
- what task should be done next
- what stage the project is currently in
Install for local development
python3 -m venv .venv
source .venv/bin/activate
pip install -e '.[dev]'
Commands
rdb init
rdb discover
rdb status
rdb next
rdb start TASK-001
rdb complete TASK-001
rdb handoff
Bootstrap workflow
- Run
rdb init - Run
rdb discover - Run
rdb status - Run
rdb next - Give the next task to Claude Code or Cline
- Commit after each completed task
Project structure
rdb init creates the following files (skips existing ones):
- context/discovery-log.md
- context/product-brief.md
- context/architecture.md
- context/decisions.md
- context/risks.md
- context/assumptions.md
- context/open-questions.md
- context/repository-context.md
- TASKS.md
- TEST_PLAN.md
- RUN_LOG.md
- PROJECT_STATE.md
- AGENT_HANDOFF.md
- .rdb/project.json
Principle
Markdown files are the source of truth. No database is required.