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 LiveTrack-X/spec-driven-ai-development --skill ai-spec-project-startgit clone --depth 1 https://github.com/LiveTrack-X/spec-driven-ai-developmentWrote 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/livetrack-x/spec-driven-ai-development/ai-spec-project-start)<a href="https://agentmods.dev/skills/livetrack-x/spec-driven-ai-development/ai-spec-project-start"><img src="https://agentmods.dev/badge/skills/livetrack-x/spec-driven-ai-development/ai-spec-project-start/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/livetrack-x/spec-driven-ai-development/ai-spec-project-start"><img src="https://agentmods.dev/badge/skills/livetrack-x/spec-driven-ai-development/ai-spec-project-start.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00067 | $0.03060 |
| Opus 5.5 | $0.00027 | $0.01224 |
| Sonnet 5.5 | $0.00013 | $0.00612 |
| Haiku 4.5 | $0.00007 | $0.00306 |
Grade A, and why
ai-spec-project-start 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 2mo 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 — 324 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI-SPEC Project Start
Create the smallest SDAD control plane that lets an owner direct AI work, separate evidence from confidence, and keep future sessions recoverable.
Reference Routing
Read references only when their trigger applies:
- Read references/runtime-contract.md before creating or changing an SDAD control surface. It owns steady-state v2 scale, execution scope, owner gates, targeted routes, evidence, and stop semantics.
- Read references/starter-templates.md when bootstrapping or repairing Standard/Full control files.
- Read references/field-patterns.md when translating lessons from another project, release/migration practice, or a reference implementation.
- Read references/implicit-rules.md when creating or repairing Core 5/control rules, or when repeated confusion or failure should become a rule, playbook, test, validator, or template.
Do not load every reference by default.
Workflow
1. Inspect Capability And Existing State
Determine whether the environment can edit the project filesystem.
Use this skill as a one-time bootstrap, install, upgrade, migration, repair, or
diagnosis entry point, not a per-session prompt. After setup, ordinary sessions
start with the installed repository adapter, then sdad-state.yaml, then
docs/INDEX.md; they invoke this skill again only for an SDAD-specific
operation named by the frontmatter trigger.
- In chat-only environments, plan and explain only. Do not install files or claim they were saved.
- In file-editing environments, inspect repository instructions, current files, version control state, source/tests, and existing SDAD surfaces before proposing changes.
- Preserve unrelated owner changes. Do not replace an existing control system merely because SDAD was requested.
- Treat review/audit requests as read-only unless the request also authorizes a change.
What ships with it
5 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.
- 2mo ago First seen · 324 lines · 67 tokens per session scan A 9649aeaeccea
ai-spec-project-start is a skill published in the GitHub repository LiveTrack-X/spec-driven-ai-development (4 stars, last pushed 1mo ago), licensed MIT. It adds 67 tokens to every session and 3,060 once invoked, about $0.0003 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.
Other skills, from other repositories
codex-issue-coordinator
Coordinates a large batch of GitHub issues through separate Codex worker threads, tested pull requests, review loops, and gated merges. Use when the user asks one Codex thread to manage several coding sessions or complete a parent issue, milestone, or issue batch.
html-doc
Generates a polished, static HTML reading view from an existing Markdown PRD or technical design. Use when a user asks to render, present, visualize, or make a PRD or design document easier for humans to read in a browser.
architecture
Writes root ARCHITECTURE.md from a fixed template: the intended system, a data flow diagram, the data model, and the rules every feature must follow. Use for a new system or a change to its boundaries, data, or deployment.
review
Runs an independent, read-only review of a code change in two parts: does it match the spec, and does it work. Tries bad input, checks security, and checks that the tests would catch a regression. Use for a diff, branch, commit, or PR.
plan
Splits a spec into small tasks, each a working slice an agent can finish in one pull request. Use when a spec is too big for one PR, or to create tracker tickets.
requirements
Writes the product requirements document (PRD) in root REQUIREMENTS.md as user stories with checkable outcomes. Use for a new product or a change to what it should do.