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/borhen68/skillengine/spec-driven-developmentnpx skills add borhen68/SkillEngine --skill spec-driven-developmentgit clone --depth 1 https://github.com/borhen68/SkillEngineWhat 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 | $0.00066 | $0.02334 |
| Opus 5 | $0.00033 | $0.01167 |
| Sonnet 5 | $0.00013 | $0.00467 |
| Haiku 4.5 | $0.00007 | $0.00233 |
Grade A, and why
spec-driven-development 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 — 246 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Spec-Driven Development
Overview
The most expensive bug is the one that never should have been built. When you write code without a spec, you're not just risking implementation errors — you're risking building the wrong thing entirely. A two-hour spec saves two weeks of rework.
The spec-driven contract: No code is written until both the agent and the human agree on what "done" looks like. The spec is the shared source of truth — it defines objectives, constraints, acceptance criteria, and boundaries. Code without a spec is expensive guessing.
Real-world impact: Teams that spec before coding ship 40% faster and have 60% fewer post-launch bugs. The time "saved" by skipping the spec is spent debugging, refactoring, and apologizing to users.
When to Use
- Starting a new project or feature
- Requirements are ambiguous or incomplete
- The change touches multiple files or modules
- You're about to make an architectural decision
- The task would take more than 30 minutes to implement
When NOT to use: Single-line fixes, typo corrections, or changes where requirements are unambiguous and self-contained.
Iron Rules
These target the specification failure modes specific to AI agents. Each is absolute.
- Never fill a gap silently. Every requirement the user didn't state but the implementation needs is an assumption — list each one and get a reaction before it hardens into code. The gaps are where the real requirements live.
- A spec must be disagreeable. If the spec only paraphrases the request back in more words, it's spec theater. A real spec contains decisions the user could object to — chosen trade-offs, named non-goals, explicit exclusions. No possible objection means no decision was made.
- Untestable criteria are not criteria. "Fast", "intuitive", "robust" cannot fail, so they cannot gate anything. Every success criterion needs a number, a command, or an observable behavior that a third party could check without asking you.
- Gates need an explicit yes. "Sounds good", "sure", or silence is not approval — re-ask, offering something concrete to disagree with. Code written past an unconfirmed gate is expensive guessing with a spec-shaped fig leaf.
- Discovering means re-specifying, not improvising. When implementation reveals the spec is wrong, stop. Update the spec, get it re-approved, then continue. Silent divergence makes the spec a lie and every future decision built on it wrong.
What ships with it
1 file 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.
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 · 246 lines · 66 tokens per session scan A 0f09a08db201
spec-driven-development is a skill published in the GitHub repository borhen68/SkillEngine (17 stars, last pushed 2mo ago), licensed MIT. It adds 66 tokens to every session and 2,334 once invoked, about $0.0003 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
code-review-and-quality
执行多维度代码审查。用于合并任何变更之前；用于审查自己、其他 agent 或人类编写的代码；用于在代码进入主分支前从多个维度评估代码质量。.
code-simplification
为清晰度简化代码。用于在不改变行为的前提下重构代码以提升清晰度；用于代码能运行但比应有状态更难阅读、维护或扩展时；用于审查已累积不必要复杂度的代码时。.
doubt-driven-development
在每个非平凡决策成立前，用全新上下文进行对抗式审查。当正确性比速度更重要、处理不熟悉代码、风险较高（生产、安全敏感逻辑、不可逆操作），或任何自信输出现在验证比之后调试更便宜时使用。.
test-driven-development
用测试驱动开发。用于实现任何逻辑、修复任何 bug，或改变任何行为。用于需要证明代码能工作、收到 bug 报告，或即将修改现有功能时。.
api-and-interface-design
指导稳定的 API 和接口设计。设计 API、模块边界或任何公共接口时使用。创建 REST 或 GraphQL endpoint、定义模块之间的类型契约，或建立前后端边界时使用。.
ci-cd-and-automation
自动化 CI/CD pipeline 设置。用于设置或修改构建和部署 pipeline 时；用于需要自动化质量门禁、在 CI 中配置 test runners，或建立部署策略时。.