Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add felipearomani/claude-shipyard/plugin install fleetWrote 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/felipearomani/claude-shipyard/planner)<a href="https://agentmods.dev/skills/felipearomani/claude-shipyard/planner"><img src="https://agentmods.dev/badge/skills/felipearomani/claude-shipyard/planner/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/felipearomani/claude-shipyard/planner"><img src="https://agentmods.dev/badge/skills/felipearomani/claude-shipyard/planner.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 | $0.00332 | $0.06459 |
| Opus 5 | $0.00166 | $0.03230 |
| Sonnet 5 | $0.00066 | $0.01292 |
| Haiku 4.5 | $0.00033 | $0.00646 |
Grade A, and why
planner 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 5d 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 — 491 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Fleet Planner
You prepare the work a fleet will execute unsupervised. The output of your work is not a handsome document: it is a dispatchable package — tasks an autonomous agent can close on its own, and that a human can verify without reopening the discussion.
The difference between planning for a person and planning for an autonomous agent is that the person asks when something is ambiguous, and the agent decides on its own — probably differently from what you wanted, and you only find out at merge time. Every ambiguity you leave becomes a decision made by somebody else, unrecorded. That is why a falsifiable acceptance criterion is not bureaucracy: it is the only channel through which you talk to whoever executes.
Setup this skill expects
- A handoff directory. Default
.fleet/<epic>/at the workspace root. If the workspace already has a convention for shared agent artifacts, use that instead — but note the circularity: a fresh coordinator session finds the plan BY the directory, so it cannot learn a non-default location from a file inside it. A non-default location has to be discoverable from a fixed point: leave a one-line pointer at.fleet/README.md("handoff lives at "), or make sure the workspace's ownCLAUDE.md/AGENTS.mdnames it, or tell the human to pass the path to the coordinator. - A tracker, if the work is going on a board. Any of Jira, GitHub Issues/Projects, Asana, Linear. Ask which one if it is not obvious.
- Nothing else. Everything below degrades gracefully when a tool is missing; where it does, the skill says what to do instead.
Where you start from
Detect this before acting, because the three inputs need different work:
- Raw idea — "I want the point-of-sale to edit the catalogue offline". Nothing exists yet: investigation, design, decomposition. Start at Phase 1.
- Finished spec — a document with the design already decided. Do not rewrite it: read it, validate it against the code (specs age), and go to Phase 3.
- Populated board with thin tasks — the tickets exist but have no acceptance criteria, no dependencies, no known traps. Run Phase 1 first, scoped per ticket — the file:line, the precedent and the trap that Phase 3.1 demands only exist if you opened the code, and enriching a ticket you never investigated produces confident criteria about code you never read. Then Phase 3, treating each ticket as a draft to enrich, preserving its key and whatever is already written.
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.
- 5d ago Changed · +20 lines 2774736f28e1
- 9d ago First seen · 471 lines · 332 tokens per session scan A 0d4a24dbb340
planner is a skill published in the GitHub repository felipearomani/claude-shipyard (1 stars, last pushed 8d ago), licensed MIT. It adds 332 tokens to every session and 6,459 once invoked, about $0.0017 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
orchestrate
Use only when the user explicitly types /orchestrate:orchestrate to decompose a large task, spawn a tree of parallel worker/subplanner/verifier subagents, and collect structured handoffs. Do not invoke autonomously.
watch
File sentinel that monitors the working directory for changes and marker comments, then auto-triggers appropriate skills. Poll-based via git diff against the last scan commit. Writes intake items for batch processing and routes marker actions through /do. Use for automatic reactions to file changes; do NOT use for…
live-preview
Mid-build visual verification loop. Takes screenshots of components during construction, not just after. Catches visual regressions and invisible features before they compound. Requires Playwright or similar screenshot tool.
marshal
Meta-orchestrator that takes any direction — broad, specific, or vague — and autonomously chains skills and context into actionable work. Gathers context from codebase, docs, and memory. Only asks the user when it genuinely cannot proceed. Single-session orchestrator.
wiki
Markdown-first knowledge base where the LLM acts as librarian. Ingests raw sources, compiles and interlinks topic files, self-maintains an index. No vector DB or embeddings required -- uses LLM-native navigation over structured markdown up to 400K words.
skill-creator
Generates Anthropic Skills with complete workflow including GitHub PR creation and local download verification.