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 skills add PANYUXIN6/pyx-skills --skill review-design-contractsgit clone --depth 1 https://github.com/PANYUXIN6/pyx-skillsWrote 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/panyuxin6/pyx-skills/review-design-contracts)<a href="https://agentmods.dev/skills/panyuxin6/pyx-skills/review-design-contracts"><img src="https://agentmods.dev/badge/skills/panyuxin6/pyx-skills/review-design-contracts/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/skills/panyuxin6/pyx-skills/review-design-contracts"><img src="https://agentmods.dev/badge/skills/panyuxin6/pyx-skills/review-design-contracts.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 85 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00150 | $0.03005 |
| Opus 5.5 | $0.00060 | $0.01202 |
| Sonnet 5.5 | $0.00030 | $0.00601 |
| Haiku 4.5 | $0.00015 | $0.00300 |
Grade A, and why
review-design-contracts scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
审查设计契约
在不进行模型投票或以 LLM 充当裁判的情况下,优先保证质量地审查设计。审查全程保持目标文档不变。
在 review.config.json 中仅使用 low 或 high 配置子代理推理强度:明确且范围小的检查用 low,复杂的契约或架构推理用 high。保持派发的任务描述不变。
选择一种操作
严格从用户明确请求中选择一种模式,不能隐式切换:
review:启动新的设计审查,且不修改目标。apply-fixes:验证并处理已接受队列、编辑目标,然后停止并交给独立审查任务。verify-fixes:独立审查已编辑目标。只有当前任务被明确要求运行verify-fixes且未编辑该目标时才能进入。
完成 apply-fixes 绝不授权同一任务进入 verify-fixes,即使已提供队列运行目录也是如此。
开始审查
- 阅读
references/review-protocol.md。 - 确认当前 Codex 任务提供 Native
spawn_agent、wait_agent和interrupt_agent。任一不可用时,在创建运行前停止;不要退回到 CLI 或 API 模型后端。 - 将
<skill-directory>解析为包含此SKILL.md的绝对目录。 - 执行
references/review-protocol.md中的权威预检查。先检查目标的 frontmatter 和直接设计链接。目标自称子设计,或委派了没有可解析路径的重要契约时,只做该文档所述的、有界的同目录和已声明索引搜索。直接关系支持的、明确且非 observed 的总设计自动纳入;否则除非未解决的外部契约对审查很重要,否则把目标视作自包含。此预检查不得调用其他技能。 - 检查目标仓库的
docs/REPO_MAP.md和docs/ARCHITECTURE.md。任一缺失时,以仓库上下文初始化模式调用$repo-map-first;相关地图说法可能过时或矛盾时,以仓库上下文验证模式调用它。保留审查目标和权威来源。观察到的地图可定位文件,但不能确立规范性权威。架构审查所需仓库上下文仍有重要歧义时,创建运行前停止并报告INSUFFICIENT_INPUT。 - 在仓库根目录运行:
node <skill-directory>/scripts/review-design.mjs prepare <design.md>
用户指定的权威项重复使用 --authority <path> 添加;预检查找到的每个明确权威项重复使用 --discovered-authority <path> 添加。不要将后者用于 observed 文档,也不要用于仅凭相邻位置、编号、时间顺序或语义相似性推断的关系。用 --retry-of <old-run-directory> 重试 FAILED 或 INVALIDATED 运行;绝不复用旧中间产物。
- 对
prepare或advance返回的每个任务描述,使用以下精确映射调用 Nativespawn_agent:
task_name ← agent_task_name
message ← spawn_message
fork_turns ← fork_turns
model ← model
reasoning_effort ← reasoning_effort
不得修改任何映射值。一个返回批次可以同时包含 L2 和互不依赖的 L3 任务;每个覆盖一个候选项或一组有界的相关候选项,分别派发。每批任务数不得超过 max_parallel_subagents。
- 等待已派发任务,不要解读其最终消息。遵循
references/review-protocol.md的超时和迟到响应协调规则;只有指定response.json存在时任务才成功。一个或多个活跃任务生成响应时运行advance;Runner 会处理完成工作、保留未完成的同级任务,并为每个空闲槽生成替代项:
node <skill-directory>/scripts/review-design.mjs advance <run-directory>
派发所有返回的重试或下一阶段描述,并持续等待它们和 waiting_for 中的 ID;不得随意等待整个批次屏障,也不得手动合并任务响应。
What ships with it
30 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 249 B
- references/adversarial-batch-result.schema.json 703 B
- references/adversarial-batch-role.md 996 B
- references/adversarial-result-legacy.schema.json 1.6 KB
- references/adversarial-result.schema.json 2.0 KB
- references/adversarial-role-legacy.md 781 B
- references/adversarial-role.md 1.1 KB
- references/architecture-fix-verification-role.md 1.1 KB
- references/architecture-merge-role.md 784 B
- references/architecture-role.md 1.1 KB
- references/architecture-shard-result.schema.json 1.3 KB
- references/architecture-shard-role.md 1.2 KB
- references/author-rebuttal-result.schema.json 772 B
- references/author-rebuttal-role.md 796 B
- references/author-response.schema.json 2.5 KB
- references/candidate-finding.schema.json 3.9 KB
- references/contract-ledger.schema.json 1.2 KB
- references/evidence-card.schema.json 2.5 KB
- references/expanded-fix-verification-result.schema.json 2.4 KB
- references/expanded-fix-verification-role.md 1.5 KB
- references/fix-verification-result.schema.json 1.7 KB
- references/fix-verification-role.md 1.0 KB
- references/human-rejection-reasons.json 1.6 KB
- references/insufficient-result.schema.json 445 B
- references/rejection-record.schema.json 1.7 KB
- references/review-protocol.md 17 KB
- references/self-consistency-result.schema.json 551 B
- references/self-consistency-role.md 1.2 KB
- review.config.json 730 B
- scripts/review-design.mjs 155 KB runs code
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.
- 1mo ago Changed · +35 tokens per session a0ae0ce92cfa
- 1mo ago Changed · +2 lines 8bdd75d8adfc
- 2mo ago First seen · 143 lines · 115 tokens per session scan A 975e02d5824f
review-design-contracts is a skill published in the GitHub repository PANYUXIN6/pyx-skills (8 stars, last pushed 17d ago), licensed MIT. It adds 150 tokens to every session and 3,005 once invoked, about $0.0006 per session on Opus 5.5. 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.
review-design-contracts compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| review-design-contractsPANYUXIN6/pyx-skills | Skill | 8 | 17d ago | MIT | A | 150 | Codex |
| autoreviewopenclaw/clawhub | Skill | 9.5k | yesterday | MIT | A | 17 | Claude Code, Codex |
| agent-teams-simplify-and-hardenLeoYeAI/openclaw-master-skills | Skill | 2.2k | 6d ago | MIT | A | 115 | Claude Code, Codex |
| omh-code-reviewrlaope/oh-my-hermes | Skill | 5.2k | yesterday | MIT | A | 48 | Claude Code, Codex |
| revdiff-planumputun/revdiff | Skill | 945 | yesterday | MIT | A | 84 | Claude Code |
| explore-unknownsdzhng/skills | Skill | 1.4k | 2d ago | MIT | A | 162 | Claude Code, Codex |
| re0-mergeLilMGenius/paperthin | Skill | 1.4k | 4d ago | MIT | A | 70 | Claude Code |