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/rsmdt/the-startup/specify-incrementalnpx skills add rsmdt/the-startup --skill specify-incrementalgit clone --depth 1 https://github.com/rsmdt/the-startupWrote 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/rsmdt/the-startup/specify-incremental)<a href="https://agentmods.dev/skills/rsmdt/the-startup/specify-incremental"><img src="https://agentmods.dev/badge/skills/rsmdt/the-startup/specify-incremental.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 | $0.00053 | $0.01179 |
| Opus 5 | $0.00026 | $0.00589 |
| Sonnet 5 | $0.00011 | $0.00236 |
| Haiku 4.5 | $0.00005 | $0.00118 |
Grade A, and why
specify-incremental 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 — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Persona
Act as an implementation planning specialist that breaks a single-feature specification into executable phases following TDD principles. Plans enable a developer to work independently without requiring clarification.
Every plan answers four questions: WHAT produces value (deliverables, not activities), IN WHAT ORDER do tasks execute (dependencies and sequencing), HOW TO VALIDATE correctness (test-first approach), and WHERE is each task specified (links to requirements.md / solution.md sections).
Spec Target: $ARGUMENTS
Interface
Task { id: string // T1.1, T1.2, T2.1, ... description: string ref?: string // Solution/Section + line range activity?: string // domain-modeling, backend-api, frontend-ui, ... parallel?: boolean prime: string // what to read before starting test: string // what to test (red) implement: string // what to build (green) validate: string // how to verify (refactor) }
Phase { number: number title: string file: string // plan/phase-N.md status: pending | in_progress | completed tasks: Task[] }
State { specDirectory = "" requirements = "" // path to requirements.md solution = "" // path to solution.md planDirectory = "" // path to plan/ subdirectory phases: Phase[] }
Constraints
Always:
- Every task produces a verifiable deliverable — not just an activity.
- All requirements.md acceptance criteria map to specific tasks.
- All solution.md components have corresponding implementation tasks.
- Dependencies are explicit with no circular dependencies.
- Every task follows TDD: Prime, Test, Implement, Validate.
- Follow template structure exactly — preserve all sections as defined.
- Wait for user confirmation before proceeding to next phase.
- Write each phase to a separate plan/phase-N.md file.
- Keep plan/README.md as the manifest with phase links and checklist.
- All tasks trace back to requirements.md or solution.md.
- Parallel tasks can actually run independently.
- Leave the
plan/README.mdphases checklist in the exact format- [ ] [Phase N: Title](phase-N.md)— this format is parsed by the implement-incremental skill for phase discovery and status tracking.
What ships with it
15 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.
- evals/evals.json 2.4 KB
- evals/fixtures/spec-single-feature/requirements.md 896 B
- evals/fixtures/spec-single-feature/solution.md 1.3 KB
- evals/fixtures/spec-validation-gaps/plan/phase-1.md 549 B
- evals/fixtures/spec-validation-gaps/plan/phase-2.md 221 B
- evals/fixtures/spec-validation-gaps/plan/README.md 418 B
- evals/fixtures/spec-validation-gaps/requirements.md 205 B
- evals/fixtures/spec-validation-gaps/solution.md 181 B
- examples/output-example.md 471 B
- examples/phase-examples.md 6.7 KB
- reference/output-format.md 418 B
- reference/task-structure.md 2.9 KB
- template.md 5.7 KB
- templates/phase.md 1.8 KB
- validation.md 4.0 KB
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 · 135 lines · 53 tokens per session scan A 3bda06e6b87d
specify-incremental is a skill published in the GitHub repository rsmdt/the-startup (511 stars, last pushed 1mo ago), licensed MIT. It adds 53 tokens to every session and 1,179 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
turnstile-core
Turnstile's non-negotiable rules -- gate order, TDD, brain-first, commit and reply conventions. Auto-loads whenever any turnstile skill is active.
turnstile-work
Implements one sprint ticket via TDD and the turnstile-coder agent, checking the brain first. Dispatched by /turnstile:work or conversate routing only.
testing-new-code
Use when adding or modifying business logic, branching code, public API, or fixing a bug in a repo using repo-hygiene — tests for the new code paths must exist before the change is committed, not "later" or "if the user asks".
job-interview-meeting-preparation
Prepare a user for a high-stakes professional meeting (interview, advisory or consulting meeting, partnership or BD meeting, or sales discovery call) with a specific stakeholder at a specific company. The skill handles input capture, web research on the company and any secondary company, stakeholder analysis from a…
comparative-landscape-brief
Produce a structured comparative briefing document for a third-party audience, across a named set of 3 to 8 entities for a named audience (investors, board, exec team, M&A committee, partners, advisory council, customer advisory board). The skill captures inputs, researches each entity, analyzes the last 90 days of…
site-to-prompt
Use when given a website URL and asked to reverse-engineer it — to produce a detailed reconstruction prompt and/or rebuild the site in HTML/CSS or React/Tailwind. Covers animated, scroll-driven, and WebGL sites.