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 OutlineDriven/odin-claude-plugin --skill spec-drivengit clone --depth 1 https://github.com/OutlineDriven/odin-claude-pluginWrote 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/outlinedriven/odin-claude-plugin/spec-driven)<a href="https://agentmods.dev/skills/outlinedriven/odin-claude-plugin/spec-driven"><img src="https://agentmods.dev/badge/skills/outlinedriven/odin-claude-plugin/spec-driven/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/outlinedriven/odin-claude-plugin/spec-driven"><img src="https://agentmods.dev/badge/skills/outlinedriven/odin-claude-plugin/spec-driven.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 38 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 | $0.00054 | $0.01102 |
| Opus 5.5 | $0.00022 | $0.00441 |
| Sonnet 5.5 | $0.00011 | $0.00220 |
| Haiku 4.5 | $0.00005 | $0.00110 |
Grade A, and why
spec-driven 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 — 52 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Spec-driven development
Contract
| Field | Bound contract |
|---|---|
| Trigger | A new project or feature is starting, requirements are unclear, the change touches multiple modules, or the task exceeds roughly 30 minutes. |
| Authority | Reversible local: files, VCS, and credentials scoped to the current working tree; no remote mutation, credential exposure, or irreversible action. |
| Side effect | Writes and commits a spec, records human-approved plan and tasks, and implements through offensive-first TDD. All writes are local; no external service calls. |
| Done | A reviewed, approved spec covering six areas is saved to the repo and the implementation follows it. |
Inputs
The user's request is mandatory. The requester must approve the spec before planning begins.
Refusal
- Human withholds approval: do not proceed to the next phase. Surface the specific objection and wait.
- Scope widens mid-implementation: stop. Return to the spec. Update it; get approval; then proceed.
- Assumption surfaces after spec is written: surface it immediately; do not proceed until the human resolves or approves it.
- No spec written or saved: the done predicate is not met. Implement nothing until the spec exists and is approved.
- Non-converged: if the human never approves, the skill ends with no implementation. State this outcome explicitly.
Procedure
- Surface assumptions first. Before writing any spec content, list every assumption being made and require explicit correction. Do not silently fill ambiguous requirements. Done when: every assumption is listed and corrected or confirmed.
- Write the spec covering six core areas: a. Objective: what is being built and why, who the user is, what success looks like. b. Commands: full executable commands with flags (build, test, lint, dev) for the stack the project actually uses. Record exact commands, not bare tool names. c. Project structure: where source lives, where tests go, where docs belong. Capture the layout that applies. d. Code style: one real snippet of the project's style beats paragraphs describing it. Include naming conventions, formatting rules, and an accepted example in the project's language. e. Testing strategy: sketch the seams where the feature will be tested. Prefer an existing seam to a new one; use the highest seam available; propose any new seam at the highest point where it can sit. Record the framework, test locations, coverage expectations, and which test level covers each concern. Confirm seams with the user before proceeding because a seam disagreement invalidates the testing strategy beneath it. f. Boundaries: three tiers — Always (run tests before commits, follow naming conventions, validate inputs); Ask first (schema changes, adding dependencies, changing CI config); Never (commit secrets, edit vendored directories, delete failing tests without approval). Done when: all six areas are written.
- Present the spec to the human. Do not advance until the human reviews and approves it. The human's approval gates every subsequent phase. Done when: the human approves the spec.
- Plan from the validated spec. Identify major components and dependencies; determine implementation order; note risks and mitigations; separate parallel from sequential work; define verification checkpoints. Done when: the plan is written.
- Present the plan to the human. Do not advance until the human reviews and approves it. Done when: the human approves the plan.
- Break the plan into discrete tasks. Each task: completable in one focused session; has explicit acceptance criteria; includes a verification step (test, build, or manual check); is ordered by dependency; changes no more than ~5 files. Done when: the task list is written.
- Present the task list to the human. Do not advance until the human reviews and approves it. Done when: the human approves the task list.
- Implement through offensive-first TDD, one task at a time. For each task: run the
tddskill's cycle — turn its acceptance criteria into attacks that must fail red for the right reason, implement until the attacks stop breaking the code, and never bend a test or fixture to reach green — then verify against the acceptance criteria before starting the next task. Load only the spec sections and source files the current task needs. Done when: every task is implemented and verified. - Update the spec first when decisions change. If the data model or scope changes, update the spec, then implement. Done when: the spec reflects the current decision.
- Commit the spec to version control alongside the code. Reference the spec in each PR. Done when: the spec is committed.
What ships with it
2 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.
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 First seen · 52 lines · 54 tokens per session scan A 443ce98336c9
spec-driven is a skill published in the GitHub repository OutlineDriven/odin-claude-plugin (37 stars, last pushed 21d ago), licensed Apache-2.0. It adds 54 tokens to every session and 1,102 once invoked, about $0.0002 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 2026-09-04.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Optimize the meaningful UI available immediately from a Next.js route on an initial load (hard navigation) or named client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route and entry point at a time; the shipped test then guards…