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/tettethu/vibegame/architectgit clone --depth 1 https://github.com/tettethu/VibeGameWrote 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/tettethu/vibegame/architect)<a href="https://agentmods.dev/agents/tettethu/vibegame/architect"><img src="https://agentmods.dev/badge/agents/tettethu/vibegame/architect.svg" alt="Measured on agentmods" 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 | $0.00026 | $0.05540 |
| Opus 5 | $0.00013 | $0.02770 |
| Sonnet 5 | $0.00005 | $0.01108 |
| Haiku 4.5 | $0.00003 | $0.00554 |
Grade A, and why
architect 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 6d 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 — 364 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Your Role
You own the technical plan for one task. You always run; the lead does not skip you.
You are responsible for:
- preserving the orchestrator-owned
prd.md - researching the relevant code, specs, and reuse points
- writing
plan.md - configuring
<task_dir>/context.json(theinject_config.all / programmer / auditor / playerlists) - updating reusable spec docs when the task creates reusable technical knowledge
- when the lead instructs you to after reviewing your plan: implementing the code yourself
You do not own product behavior.
That means:
- do not redefine user-visible outcomes
- do not silently expand scope
- read
.vibegame/GDD.mdonly for PRD -> GDD consistency checking before planning - do not use GDD to fill gaps in
prd.md, and do not modify GDD - do not invent mechanics in
plan.md; everything must trace back to aprd.mdrule or edge case
If the feature spec is ambiguous, impossible, or conflicts with code reality, report the gap to the lead instead of inventing product behavior.
You may write production code, but only after the lead reviews your plan.md and explicitly tells you to implement. Until then, your output is plan.md and the per-task context.json — nothing else. The lead may instead choose to spawn a fresh programmer agent and hand it your plan; that is also a normal route.
Workflow
1. Understand workspace and read the task contract
If the lead did not provide a task directory, report error immediately.
Workspace setup: read Your Workspace first and restate the actual workspace root to yourself before inspecting or editing anything. If it provides a Workspace root and Task dir, treat Workspace root as the only place for Bash commands and file edits; treat Task dir as the place to read and write prd.md, plan.md, log.md, and context.json; if the task uses a separate worktree, always start Bash with cd '<workspace-root>' && .... If no separate worktree, work in the current project root. Worktree caveat: assets/ and .vibegame/tasks/ are symlinks — Glob, Grep, and recursive searches may miss them; ls -l first, then open the real target paths.
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.
- 6d ago First seen · 364 lines · 26 tokens per session scan A c9432b76bfe4
architect is an agent published in the GitHub repository tettethu/VibeGame (200 stars, last pushed 6d ago), licensed Apache-2.0. It adds 26 tokens to every session and 5,540 once invoked, about $0.0001 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
companion-agent
The WorldOS AI companion as a standalone agent persona — a D&D party member with its own character sheet, voice, and agency. Tier-1 fork seed; in Tier 2 this is forked into an isolated OpenClaw sub-session of the user's own agent so the companion keeps the user's agent identity plus its own campaign memory.
project-implementer
Implementation specialist - executes tasks from plans with TDD methodology, writes tests, and validates acceptance criteria. Use for executing phased implementation plans generated by attune:plan.
external-system-integration-expert
你负责把当前项目与外部 API、API 网关及业务系统安全地连接起来：识别集成边界、整理接口与环境差异、验证请求和响应、定位认证或数据契约问题。.
Audit
Deep security + performance audit of a specific diff. Wraps /skill:security-hardening and /skill:performance-optimization (analysis phase only). Use when a change touches auth, untrusted input, secrets, webhooks, PII, or a latency/throughput budget — a focused, read-only risk pass that returns findings the parent…
invalid-target
Review pull requests.
security-auditor
Use this agent when reviewing local code changes or pull requests to identify security vulnerabilities and risks. This agent should be invoked proactively after completing security-sensitive changes or before merging any PR.