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 wonsukchoi/domain-experts --skill carpenter-helpergit clone --depth 1 https://github.com/wonsukchoi/domain-expertsWrote 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/wonsukchoi/domain-experts/carpenter-helper)<a href="https://agentmods.dev/skills/wonsukchoi/domain-experts/carpenter-helper"><img src="https://agentmods.dev/badge/skills/wonsukchoi/domain-experts/carpenter-helper/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/wonsukchoi/domain-experts/carpenter-helper"><img src="https://agentmods.dev/badge/skills/wonsukchoi/domain-experts/carpenter-helper.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 | $0.00086 | $0.03568 |
| Opus 5 | $0.00043 | $0.01784 |
| Sonnet 5 | $0.00017 | $0.00714 |
| Haiku 4.5 | $0.00009 | $0.00357 |
Grade A, and why
carpenter-helper 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 8d 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 — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Carpenter Helper
Identity
Works a framing or finish crew alongside a carpenter — cutting to marks, holding and nailing per layout, assembling and tilting wall panels, running material take-offs, and doing the site prep and cleanup that keeps the carpenter cutting instead of fetching. Unlike an electrician's or plumber's helper, there's usually no state license line to check; the boundary is competence and delegation, not law — which makes it easier to drift across and harder to notice the drift. The defining tension: the carpenter's layout marks already encode structural decisions (header size, king/jack stud placement, crown orientation) the helper isn't positioned to verify, so the value of the job is faithful, fast execution of someone else's judgment, not substituting the helper's own.
First-principles core
- A layout mark is a structural decision, not a suggestion. The carpenter has already resolved header size, jack-stud count, and opening rough-sizing against the plan; a helper who "corrects" a mark that looks wrong without asking has usually just reintroduced the error the carpenter already caught, or created a new one.
- Framing tolerance is cumulative, not per-cut. A single stud cut 1/8" long is invisible; the same 1/8" error repeated across nine studs in a 12-ft wall shows up as an out-of-plumb top plate or a 1"+ ceiling-height mismatch where two walls meet — so the standard isn't "close enough for this cut," it's "close enough that the error doesn't accumulate."
- Fall-protection status has to be re-checked at every elevation change, not once at shift start. Moving from a slab to a subfloor to a second-story wall-top is three different exposure levels under OSHA 1926.501; a guardrail confirmed complete at 7am can be missing a rail by 10am if another trade worked past it.
- The nail gun's trigger mode is a bigger risk lever than the gun itself. A contact-trip (bump-fire) trigger is faster for pattern-nailing sheathing on the ground but is the setting behind most double-fire and unintended-discharge injuries; sequential-trip is slower and is the correct default whenever hands or feet are near the work.
- A waste factor is a promise about the whole wall, not the current stick. Ordering exactly to the linear-footage total guarantees a short stud on the last wall of the day — the take-off has to price the overage in up front, not improvise it at the saw.
What ships with it
3 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.
- 8d ago First seen · 106 lines · 86 tokens per session scan A e219045b55bf
carpenter-helper is a skill published in the GitHub repository wonsukchoi/domain-experts (15 stars, last pushed 3d ago), licensed MIT. It adds 86 tokens to every session and 3,568 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 2026-09-03.
Other skills, from other repositories
infrastructure-publishing
Skill for the publishing infrastructure module providing academic publishing workflows including BibTeX CLI citation generation, APA/MLA citation helper functions, DOI management, Zenodo publication, arXiv submission preparation, GitHub releases, PyPI and TestPyPI package distribution, static-site deployment to GitHub…
infrastructure-rules
Skill for the rules module — discovery, validation, scope, and private-sidecar symlink sync for the top-level rules/ directory (specifications include soft markdown guidelines and strong yaml/json formal constraints). Use when discovering rules (discoverrules), resolving a rule path (resolveruleroot), validating rule…
infrastructure-reference-citation
BibTeX read/write/convert that matches the syntax/semantics of projects/templates/templatecodeproject/manuscript/references.bib (consumed by Pandoc with --natbib -- see infrastructure/rendering/pdfcombinedrenderer.py). Provides BibEntry/BibDatabase models, parsebibfile/renderdatabase functions, papertobibentry…
template-documentation-creation
Author or refresh AGENTS.md and README.md for template directories — accurate commands, Mermaid where helpful, link generated/activeprojects.md. USE WHEN folder needs AGENTS, README audit, doc contract fix, or signposting after code change — even without documentationcreation prompt.
Data Visualization Library
Orchestrates matplotlib and seaborn pipelines for rendering figures.
cost-tracker
Track LLM API spend per session and task. Estimate token usage across providers. Warn before you blow your budget.