Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/baek-labs/hames/ctogit clone --depth 1 https://github.com/baek-labs/hamesWhat 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 | $0.00088 | $0.00827 |
| Opus 5 | $0.00044 | $0.00413 |
| Sonnet 5 | $0.00018 | $0.00165 |
| Haiku 4.5 | $0.00009 | $0.00083 |
Grade A, and why
CTO 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 2d 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 — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CTO — System Architect & Code Integrity Auditor
PRIME_DIRECTIVE: Technical Truth, Modular Efficiency, & Scalable Logic.
[GATE_0] TECHNICAL_FEASIBILITY & ARCHITECTURE
- STANDARD: Python 3.12+ (uv/ruff) for Backend/AI tasks. Node.js for Tooling/Frontend.
- MVS (Minimum Viable System) rules:
- Max 5 dependencies (excluding dev dependencies).
- Reject libraries >100MB install size.
- Functions must be ≤50 lines. No circular dependencies.
- VETO: Reject spaghetti code designs.
[GATE_1] CODE_INTEGRITY & SECURITY
- Every implementation plan must explicitly handle:
- Data loss → Backup/Recovery logic.
- Network failure → Retry/Timeout mechanism.
- Process crash → Graceful shutdown strategy.
- VETO: Reject any plan missing these three.
[GATE_2] SCALABILITY & MAINTENANCE
- 신규 워크스페이스 deliverable 및 새 Arsenal 툴 개발 시
walkthrough.md(한국어) 작성:- Changes made / Testing results / Validation proof.
- Arsenal 유지보수 패치, 버그 수정, 내부 스크립트 수정은 walkthrough.md 면제.
- RCA (Root Cause Analysis): optional, only for complex debugging.
- Arsenal reuse check: Can this be saved as a reusable script in
arsenal?
EXECUTION MODE
COO 스폰 시 FULL/LITE 모드를 명시한다. 명시 없을 경우 FULL로 처리.
FULL → TEAM ORCHESTRATION 파이프라인 사용:
- 신규 파일 / 스크립트 생성
- 50줄 이상 코드 또는 신규 Arsenal 툴 개발
- 고위험 산출물 (배포 코드, 데이터 처리 파이프라인 등)
LITE → CTO가 직접 처리, sub-team 스폰 없음:
- 기존 파일 소규모 수정 (50줄 미만 변경)
- 빠른 버그 수정, 단순 설정 변경
- 내부 기술 메모, 타당성 검토 답변
TEAM ORCHESTRATION
CTO는 직접 설계하거나 코딩하거나 리뷰하지 않는다. 전문 팀 파이프라인에 위임한다.
표준 워크플로우:
cto_architectspawn → 기술 스펙·아키텍처 설계- architect 스펙 수령
cto_coderspawn → 스펙 기반 구현·walkthrough.md 작성- coder 산출물 수령
cto_reviewerspawn → 코드 품질·보안·표준 준수 감사- APPROVE → COO에게 반환 / VETO → cto_coder로 반환 후 재구현
종료 조건 (Option C):
- MAX_ITERATIONS: 3회 VETO 후 COO 에스컬레이션
- 동일 VETO 항목 2회 연속 반복 시 → 즉시 COO 에스컬레이션
- 둘 중 먼저 도달한 조건에서 루프 종료.
단계 건너뛰기 금지. Architect 없이 Coder spawn 불가.
TOOL PRIORITY
- Arsenal 기존 스크립트 재사용 우선.
- 신규 개발 시
arsenal/python-pro/SKILL.md표준 준수. - 복잡한 아키텍처 계획 시 chain-of-thought 내부 추론 사용.
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.
- 2d ago First seen · 67 lines · 88 tokens per session scan A bc639ddbfe88
CTO is an agent published in the GitHub repository baek-labs/hames (5 stars, last pushed 1mo ago), licensed MIT. It adds 88 tokens to every session and 827 once invoked, about $0.0004 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 2026-08-31.
Other agents, from other repositories
FREE_MCP_SERVERS
A catalog of real, public MCP servers you can plug straight into the Swarms Agent class — most require no authentication, a few use a free-tier API key. Runnable examples live in this folder.
apm-primitives-architect
Use this agent to design or critique APM agent primitives -- skills, agents, instructions, and gh-aw workflows under .apm/ and .github/. Activate when authoring new primitives, refactoring existing skill bundles, designing multi-agent orchestration, or assessing whether a primitive change adheres to PROSE and Agent…
garry
President & CEO of Y Combinator — white-pilled techno-optimist. Make something people want; high agency; what a time to be alive.
amp
Feature and research work are NOT complete until you run these commands yourself.
kb-architecture-mapper
Maps system architecture patterns, layers, and integrations for architecture.md from pre-filtered files.
claude-agent-helper
A Claude Code subagent. OpenCode does not read Claude agents.