docs: add Operator OS case study + 90-second demo plan#48
Conversation
Public, sanitized write-up of the multi-agent control plane over a 129-repo portfolio: truth (auditor) -> shared state (bridge-db) -> events (notification-hub) -> desktop command center -> cost/health overlays. Metrics drawn from a real portfolio-truth-latest.json snapshot (schema 0.5.0). DEMO-PLAN.md scripts a 90s recording across the five PortfolioCommandCenter tabs. Sanitized: no employer identifiers, emails, absolute paths, handles, IPs, secrets, or individual portfolio repo names.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 8747bf04b2
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| 1. **Refresh the producer artifacts** so the snapshot is today's: | ||
| ```sh | ||
| # in the auditor repo — flags FIRST, then username, run via python -m | ||
| python -m src.cli --portfolio-truth --portfolio-truth-include-security <user> |
There was a problem hiding this comment.
In the documented pre-record setup, this command cannot refresh the snapshot because the CLI does not define --portfolio-truth-include-security (the portfolio-truth parser only has --portfolio-truth and --portfolio-truth-include-release-count in src/cli.py), so argparse exits with an unrecognized-argument error before generating portfolio-truth-latest.json. Anyone following the demo plan verbatim will be blocked at the first setup step; use the supported portfolio-truth invocation or add the flag before documenting it.
Useful? React with 👍 / 👎.
1 participant
What this adds
Two public, sanitized docs describing the Operator OS — the multi-agent control plane over a 129-repo portfolio:
git logacross 100+ repos can't grade real state), the architecture (truth → state → events → surfacewith a mermaid diagram + component table), real metrics from aportfolio-truth-latest.jsonsnapshot, and the Claude Code + Codex coordination model (lanes, handoffs, push policy).Real metrics (truth schema 0.5.0)
129 projects · 22 recent / 90 active / 5 stale / 12 archived · 49 repos with open high/critical security alerts · agent attribution Claude Code 53 / Codex 23 / GPT 12 / unknown 41.
Notes