Three months of running a consultancy out of a git repo our coding agents read

Description and use

BKS‑Lab runs a consultancy using a git‑only workflow with AI coding agents, achieving scalable memory, team coordination, and per‑company isolation.

Problem

Managing multiple clients, projects, and team workflows with a coding agent that relies on a single git repository for memory and configuration.

Outcome

After three months, BKS‑Lab successfully runs a consultancy using a git‑only workflow with coding agents, achieving scalable memory management, team coordination, and per‑company isolation.

Workflow

• 1. Copy the setup prompt from https://bks-lab.github.io/open-bridge/#get-started and paste into Claude Code, Codex or Copilot CLI.
• 2. Agent checks tools, shows plan, and creates a private repo.
• 3. Onboarding walks through identity, repos, clients, and generates a first briefing from the setup.
• 4. Daily work logs are written as individual rows immediately after each task.
• 5. Each task resides in its own folder with a STATUS.md header and free text.
• 6. The board is generated automatically from task folders.
• 7. Private data lives on a user branch; template on main; pre‑push hook blocks pushing private branch to public remote.
• 8. CI enforces byte budget for always‑loaded part; larger data is indexed and fetched on demand.
• 9. A plain table maps everyday questions to files and validates links.

Evidence

open-bridge is a git repo of markdown and YAML. Claude Code, Codex or Copilot CLI read it at the start of every session

Source

The daily log is the real memory. Every unit of work gets one row: date and time, a glyph, the project, what happened.

Source

Private data and the shared template never touch. My data lives on a private user branch, the template lives on main.

Source