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 fledgeling-co/fledgeling-plugins --skill harbourmastergit clone --depth 1 https://github.com/fledgeling-co/fledgeling-pluginsWrote 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/fledgeling-co/fledgeling-plugins/harbourmaster)<a href="https://agentmods.dev/skills/fledgeling-co/fledgeling-plugins/harbourmaster"><img src="https://agentmods.dev/badge/skills/fledgeling-co/fledgeling-plugins/harbourmaster.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 5.1 | $0.00302 | $0.03119 |
| Opus 5 | $0.00151 | $0.01559 |
| Sonnet 5 | $0.00060 | $0.00624 |
| Haiku 4.5 | $0.00030 | $0.00312 |
Grade A, and why
harbourmaster 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 6d 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 — 230 lines — stays where its author put it; the contents beside it link to each section on GitHub.
harbourmaster
A harbourmaster does not sail the ships. They decide which berth a vessel takes, whether it may enter yet, and which ones wait outside — and they keep the register of what is in port.
This skill answers two questions, and the second only matters because of the first:
- Which plane should this work run on? Five are available and they spend different resources. The wrong plane is not slower; it is billed to a resource that had none left.
- May it start on this Mac yet? Only the local plane needs governing, because it is the only one whose scarcity this machine feels.
The measured problem it exists for: on 2026-08-22 this Mac carried a load average
of 830 across 16 cores — 76 claude processes, 73 node, 19 rustc — while
ship-fleet was starting runners from a hard-coded count of 8 with no reading of
the machine at all. Nothing was wrong with any single decision. Nothing was
counting them together.
Running as a Gemini model? Read gemini.md in this directory first, then follow this file with the overrides it names. It turns the five-plane table into a probe ledger with a fraction to report, reads exit 75 as a retry ceiling rather than a queue to loop on, and requires every pressure, berth and thermal number to arrive with the command that produced it. Other models skip it.
Route first
Ask what each plane spends. Take the plane whose resource is cheapest right now, not the one that is nearest to hand.
| Plane | Spends | Reach for it when |
|---|---|---|
This Mac via governor-run |
CPU, RAM, disk — the thing that is scarce | Building, testing, running anything that must touch this working tree |
anvil-errand container on the node |
Another machine entirely; nothing here | Long, self-contained, CPU-heavy work that needs no local state — benchmark runs, big builds, batch jobs |
proctor instrument session |
A machine-wide foreground turn, and little CPU | Driving or verifying a native macOS app — the accessibility tree, capture trustworthiness, geometry assertions |
defer to another model or CLI |
Another vendor's plan headroom | Judgment, verdicts, second opinions, out-of-family review, completeness critique — unless the caller is running tiered, which turns this plane off |
| Claude subagents / workflows | This session's rate limit and context | Wide reading, search and investigation across many files |
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.
- gemini.md 16 KB
- references/admission.md 6.2 KB
- references/evidence.md 5.0 KB
- references/integration.md 3.0 KB
- references/routing.md 3.3 KB
- references/thermal.md 4.3 KB
- scripts/.gitignore 13 B
- scripts/berths.py 9.7 KB runs code
- scripts/demote.py 7.2 KB runs code
- scripts/governor-run 7.9 KB
- scripts/install.sh 6.0 KB runs code
- scripts/ledger.py 6.3 KB runs code
- scripts/pressure.py 13 KB runs code
- scripts/selftest.sh 5.5 KB runs code
- scripts/thermal.py 13 KB runs code
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.
- 6d ago First seen · 230 lines · 302 tokens per session scan A f1d68d877896
harbourmaster is a skill published in the GitHub repository fledgeling-co/fledgeling-plugins (2 stars, last pushed today), licensed MIT. It adds 302 tokens to every session and 3,119 once invoked, about $0.0015 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
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…