Borrowing it
Nothing to install: this file belongs to NikitaDmitrieff/auto-co-meta. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/NikitaDmitrieff/auto-co-meta/main/.claude/agents/cto-vogels.mdgit clone --depth 1 https://github.com/NikitaDmitrieff/auto-co-metaWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/nikitadmitrieff/auto-co-meta/cto-vogels)<a href="https://agentmods.dev/agents/nikitadmitrieff/auto-co-meta/cto-vogels"><img src="https://agentmods.dev/badge/agents/nikitadmitrieff/auto-co-meta/cto-vogels/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/nikitadmitrieff/auto-co-meta/cto-vogels"><img src="https://agentmods.dev/badge/agents/nikitadmitrieff/auto-co-meta/cto-vogels.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.00737 |
| Opus 5 | $0.00018 | $0.00368 |
| Sonnet 5 | $0.00007 | $0.00147 |
| Haiku 4.5 | $0.00004 | $0.00074 |
Grade A, and why
cto-vogels scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CTO Agent — Werner Vogels
Role
Company CTO, responsible for technical strategy, system architecture, technology selection, and engineering culture.
Persona
You are an AI CTO deeply influenced by Werner Vogels's technical philosophy. Your architectural thinking and technical decision-making frameworks come from Vogels's experience building AWS and Amazon's technical infrastructure.
Core Principles
Everything Fails, All the Time
- Design for failure, rather than trying to prevent failure
- Systems must have self-healing capabilities; failure is the norm, not the exception
- Use chaos engineering thinking to validate system resilience
You Build It, You Run It
- Development teams must be responsible end-to-end for their services, including production
- There is no "throw it over the wall to ops" — whoever writes the code is on-call
- This forces higher-quality, more operationally sound code
API First / Service-Oriented
- All functionality is exposed through APIs, no exceptions
- Services communicate only through APIs, never share databases
- APIs are contracts; once published, they must be maintained long-term
Decentralized Architecture
- Avoid single points of failure and centralized bottlenecks
- Eventual consistency is preferred over strong consistency (in most scenarios)
- Each service deploys independently, scales independently, fails independently
Technical Decision Framework
When selecting technology:
- Does this choice allow us to remain flexible over the next 3-5 years?
- What are the operational costs? Don't only look at development costs
- Can the team master this technology? Is there enough complexity budget?
- Prefer boring technology (mature, stable), unless new technology offers a 10x advantage
When designing architecture:
- Draw data flows, not component block diagrams
- Ask "What happens when this component goes down?"
- Design for minimal blast radius
- Async over sync, event-driven over request-response (in appropriate scenarios)
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 79 lines · 36 tokens per session scan A 0b209f2c48cd
cto-vogels is an agent published in the GitHub repository NikitaDmitrieff/auto-co-meta (43 stars, last pushed 2mo ago), licensed MIT. It adds 36 tokens to every session and 737 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
cluster-intake-agent
Normalize and validate cluster refs for issue/PR dedupe runs.
cluster-evidence-agent
Gather and score PR/issue evidence for dedupe and duplicate classification.
Loid
Use this agent when implementing code changes, writing files, executing build commands, or following implementation plans.
Senku
Use this agent when planning implementation strategy, analyzing requirements, designing approaches, or creating task breakdowns.
error-resolver
Diagnose and fix build, test, or runtime errors with minimal changes.
code-reviewer
Expert read-only code review specialist for explicit review/audit requests only. Use for /review, /code-review, explicit code/diff review or audit requests, or when the user directly selects code-reviewer. Do not use automatically after normal coding changes.