What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
After three months using a Git repository as working memory for consultancy work, Michael Boiman says the most useful habits were simple: record work as it happens, give each task its own folder, and keep a human in charge of consequential actions. The system helped his team organize context for coding agents, but the activity counts he reports are not evidence of time saved or better client outcomes.
What the repository does
Boiman describes open-bridge as “a git repo of markdown and YAML.” Its files record clients, repositories, open tasks, and recent work. At the start of a session, Claude Code, Codex, or Copilot CLI reads relevant files so it has operational context before responding. The project documentation describes a registry, task folders, skills, and standing orders. (Boiman’s account; project README)
This is not an autonomous consultancy. The repository gives an agent information to work with; it does not take responsibility for business decisions or actions. Boiman says he keeps email as drafts, makes merge decisions himself, and requires approval before deletion. The README also characterizes its safety rules and status conventions as conventions, not as an operating-system sandbox.
Habits that kept the context useful
Write the log when work lands
Boiman calls the daily log “the real memory.” Each entry records a date and time, a glyph, a project, and what happened. His rule is to write the row when work lands rather than reconstructing the day later. A morning briefing can then draw on recent entries to help the agent—and the operator—pick up where work stopped.
#1 Best Overall
Give each task its own folder
Each task has a folder containing a STATUS.md file with a small YAML header and free text. The task board is generated from those folders instead of being maintained as a separate manual record. That reduces the chance that the board and underlying task files disagree. As Boiman puts it, “One folder per task beats one big notes file.”
Separate personal and shared material deliberately
In Boiman’s described setup, personal data lives on a private user branch, while the template lives on main. He says their paths are disjoint and a pre-push hook refuses to push the private branch to a public remote. This is a configuration choice, not a general security guarantee.
For work involving several companies, the account describes one instance per company, with client data kept out of other clients’ repositories. Shared company skills, accounts, projects, and rules sit in a separate overlay repository that instances subscribe to. The goal is to share reusable team material without putting different clients’ records in the same repository.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →What changed as the system grew
Always-loaded context needed a budget
The files loaded before an agent’s first answer grew beyond what the team intended. Boiman says the response was to set a byte budget for always-loaded context and make CI fail if that budget is exceeded. Material that does not need to load every time became an index: one line per client or repository, with fuller information fetched only when relevant.
People needed a map to the information
Team members repeatedly asked where particular information lived. The team added a plain table mapping everyday questions to the file that answers them, along with a check that the table’s links resolve. The lesson is practical: a repository can contain useful notes and still be hard to use unless people have a dependable way to find them.
The work model needed to match the operator
The initial worked example centered on an agency with two clients. Boiman says the most-used instance instead belonged to one person wearing several hats: consultancy partner, freelancer, household manager, and home-server operator. The project then added a second example for that broader mix. A useful information structure should reflect the actual shape of someone’s work, not just the first example chosen to explain it.
What Boiman’s activity counts do—and do not—show
For his own instance, Boiman reports the following counts, measured on September 24, 2026:
| Activity | Reported count | Qualification |
|---|---|---|
| Work-log rows | 4,072 | Since July 4, 2026; Boiman’s instance |
| Closed tasks | 164 | Boiman’s instance |
| Commits | 2,426 | Since June 20, 2026; Boiman’s instance |
These are measures of recorded activity, not measures of productivity. They do not establish time saved, improved work quality, client satisfaction, or a causal benefit from using the repository. The README says the project does not yet publish measured benefit numbers.
Best Value
Who might find this approach useful
The README’s intended audience is people managing multiple clients, repositories, or roles. If your work has those kinds of context boundaries, plain files in a private repository may offer a way to keep records versioned and readable without running a database or hosted service. Someone working in a single repository may find a simpler CLAUDE.md instruction file sufficient.
The trade-off is structure and upkeep: task folders, logs, indexes, context budgets, and checks all need to stay current. The available accounts do not quantify setup or maintenance costs or compare this method head-to-head with other systems. The sources also do not show that repository-based working produces better consultancy outcomes.
Confidentiality and safety still depend on your setup
A private repository and a hook are not substitutes for assessing client obligations. Boiman’s described branch arrangement is one team’s configuration; it does not establish that Git automatically meets confidentiality, regulatory, contractual, access-control, backup, or secret-handling requirements. The README explicitly says the project’s safeguards are not an OS-level sandbox.
Free tools Windows power users keep installed
One-click scans. No signup required.
Before placing client information in any agent-readable repository, decide what data may be stored there, who can access it, how secrets are excluded, and what happens to copies and backups. Keep approval gates for actions with external or destructive effects. Those boundaries matter because an agent’s access to context does not transfer accountability away from the person using it.
What three months established
Boiman’s account offers a concrete operating pattern: a repository can act as shared working memory when its logs are timely, tasks have clear records, and context is retrieved selectively. It also shows the maintenance problems that emerge as files accumulate—oversized startup context, hard-to-find notes, and examples that do not fit every operator. The reported counts document use in one instance; they do not settle whether the approach outperforms alternatives.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
