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/shipwrights/core/shipwrights-specnpx skills add shipwrights/core --skill shipwrights-specgit clone --depth 1 https://github.com/shipwrights/coreWrote 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/shipwrights/core/shipwrights-spec)<a href="https://agentmods.dev/skills/shipwrights/core/shipwrights-spec"><img src="https://agentmods.dev/badge/skills/shipwrights/core/shipwrights-spec.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 | $0.00075 | $0.01826 |
| Opus 5 | $0.00037 | $0.00913 |
| Sonnet 5 | $0.00015 | $0.00365 |
| Haiku 4.5 | $0.00007 | $0.00183 |
Grade A, and why
shipwrights-spec 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 — 174 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/shipwrights-spec — draft a spec for human approval
You are the orchestrator for a single task. Unlike /shipwrights-epic, which drives a refined backlog item through the full pipeline, this skill takes an ad-hoc task description and walks 4 stages of analysis before halting at a human approval gate.
The goal is a plan document the user can read and approve before any source file is written.
What you produce
A single plan document at <spec.output_dir>/<id>-<slug>.md (default docs/backlog/specs/). The user reviews it, then runs one of:
/shipwrights-spec-approve <id>— proceed to build/shipwrights-spec-revise <id> <note>— re-run analyze + plan with the note/shipwrights-spec-cancel <id>— tombstone the draft
If the user invoked /shipwrights-spec with --auto, you skip the gate and proceed straight to build after writing the plan.
Inputs
Either:
- A task description string ("add a forgot-password flow"). Allocate a new id, write the input file under
.shipwrights/specs/<id>.input.md, then run the 4 stages. - An existing id (
S-<yyyy-mm-dd>-<hash>). Read the input file, check the plan's status. Ifdraftedandrevisions > 0, you're being asked to redraft after a revision. Ifdraftedwithout revisions, the previous run was interrupted — resume from wherever the manifest / plan are missing.
The 4 stages
Stage 1: Discover (read-only)
Read enough of the codebase to ground the spec in reality. Honour spec.context_depth:
shallow— read 5-10 high-signal files (entry points,package.json,CLAUDE.md/AGENTS.md, the closest existing module to the task).medium(default) — read 20-30 files; spread across the layers the task touches.deep— read 50+ files; everywhere the task could plausibly affect.
Emit a manifest at .shipwrights/specs/<id>.manifest.json listing every file read (with line ranges) and key observations (conventions, frameworks, primitives worth reusing).
Hard rule: subsequent stages may only cite paths that appear in the manifest. Treat the manifest as the authority. If you discover later that you need to cite a file you didn't read, stop and re-run discover rather than guessing.
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 · 174 lines · 75 tokens per session scan A 8485184bd00f
shipwrights-spec is a skill published in the GitHub repository shipwrights/core (2 stars, last pushed 1mo ago), licensed MIT. It adds 75 tokens to every session and 1,826 once invoked, about $0.0004 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 2026-08-31.
Other skills, from other repositories
check-pr
Read-only inspection of a single GitHub PR lifecycle — checks CI, review threads, description sync, and mergeability, and returns PASS or FAIL with per-gate findings. Never invokes the merge button. Use when verifying a PR is ready to merge, polling lifecycle progress, checking mergeability, or babysitting a GitHub PR…
babysit-pr
Author-side PR lifecycle babysitter and companion to review-pr. Use when the user wants to tend an existing GitHub PR through CI, review threads, description sync, mergeability, auto-fixes, or asks to babysit a PR with or without an existing manifest.
just-define
Goal-based Manifest builder. Encodes shared understanding into a verifiable Manifest — Deliverables, Acceptance Criteria, Global Invariants — deciding for itself how to interview. Use when the user asks to just define it, spec it lean, or make a manifest with minimal process.
figure-out
Figure things out together — any topic, problem, or idea. Presses relentlessly until shared understanding is reached. Use when understanding is the deliverable rather than a preamble to acting, when figuring it out is the goal, or when the user asks to think through a decision, dig deeper, press an assumption…
review-writing
Review any prose against this project's writing standards — a README, a spec, a design doc, a blog post, marketing copy, or a draft you paste in. Detects whether the text is documentation or human-voiced writing, applies that register's rules plus the rules holding in both, and reports graded findings with the fix.…
walk-pr
Walk through a PR or large diff together — your own or someone else's — one sub-changeset at a time and one review topic at a time within it. Use when reviewing a substantive PR collaboratively, walking a large refactor, or when the user asks to walk a PR, walk a diff, explain a PR, review collaboratively, or review a…