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 skills/extra-org/extra/architecture-reviewnpx skills add extra-org/extra --skill architecture-reviewgit clone --depth 1 https://github.com/extra-org/extraWrote 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/skills/extra-org/extra/architecture-review)<a href="https://agentmods.dev/skills/extra-org/extra/architecture-review"><img src="https://agentmods.dev/badge/skills/extra-org/extra/architecture-review.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 | $0.00036 | $0.00990 |
| Opus 5 | $0.00018 | $0.00495 |
| Sonnet 5 | $0.00007 | $0.00198 |
| Haiku 4.5 | $0.00004 | $0.00099 |
Grade A, and why
architecture-review 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 4d 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 — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill: Architecture Review
Purpose
Guard the project's architecture invariants. Use this to evaluate any change that affects layers, boundaries, lifecycles, or contracts — and to decide whether an ADR is required.
When to Use This Skill
- A change touches more than one layer or moves logic between layers.
- A change affects the runtime lifecycle, the YAML→graph pipeline, prompts, plugin contracts, or tool/MCP integration.
- Reviewing whether a design respects the project's invariants.
- Before approving anything that alters a public contract.
Files to Read First
AGENTS.md§3 (non-negotiable rules) and §4 (layout).docs/ARCHITECTURE.mdanddocs/RUNTIME_LIFECYCLE.md.- All ADRs in
docs/adr/. - The specific layer doc(s) the change touches.
Core Principles (project architecture invariants)
These are binding. A change that violates one must be redesigned or justified by a new/updated ADR.
- YAML is declarative, not executable logic. No code execution from spec values; routing conditions are declarative.
- Validate before compile. YAML is validated before it is compiled.
- Runtime never executes raw YAML dictionaries. It operates only on typed compiled models.
- Validated config compiles into typed internal models (
CompiledAgentGraph). RuntimeEngineis created once at startup; the compiled graph is immutable and shared read-only.ExecutionContextis created per request and holds all request-scoped state. No request state on the engine/graph.- Prompt templates may be cached; rendered values are resolved per request. No global cache of rendered prompts.
- Client-specific auth/context/business logic lives in plugin boundary, never in the runtime.
- Tool permissions are enforced outside prompts, at the tool/data layer; injected params can't be overridden by the model.
- Agents declare needs; the runtime resolves and enforces them.
- Plugin contracts stay generic — no client-specific fields baked into it.
- MCP/tool integrations go through adapters, not directly into core logic.
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.
- 4d ago First seen · 104 lines · 36 tokens per session scan A c6a64d000388
architecture-review is a skill published in the GitHub repository extra-org/extra (109 stars, last pushed yesterday), licensed MIT. It adds 36 tokens to every session and 990 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 skills, from other repositories
deck-course-module
暖纸背景 + Playfair, 左侧学习目标常驻, 含 MCQ 自测页.
skill-intent-contract
Use when starting a complex or ambiguous task that risks scope drift.
argent-tv-interact
Control and inspect TV apps via argent — Apple TV (tvOS), Android TV (leanback), and Amazon Fire TV (Vega). Boot the target, read focus, navigate with the D-pad remote, type, screenshot, and on Vega debug the JS runtime (evaluate, console logs, network inspector). Use when a task targets a TV (runtimeKind "tv", or…
bernstein-approve
Review and approve/reject pending tasks or plans in Bernstein. Use when the user asks about approvals, wants to review agent work, or needs to approve/reject a plan before execution begins.
bernstein-quality
Show quality metrics for Bernstein runs - success rates per model, lint/test pass rates, completion time distributions. Use when the user asks about quality, reliability, which model performs best, or pass rates.
review-offered-task
Review a task that has been offered to you and decide whether to accept or reject it.