Lathe is a tool that uses language models to create hands-on, multi-part technical tutorials and lets people study them in a local interface. It is intended for learners who want to follow tutorials themselves, while coding agents can generate, verify, extend, and manage those tutorials through its skills and command-line tool.
Borrowing it
Nothing to install: this file belongs to devenjarvis/lathe. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/devenjarvis/lathe/main/.claude/skills/lathe-verify/SKILL.mdgit clone --depth 1 https://github.com/devenjarvis/latheWrote 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/devenjarvis/lathe/lathe-verify)<a href="https://agentmods.dev/skills/devenjarvis/lathe/lathe-verify"><img src="https://agentmods.dev/badge/skills/devenjarvis/lathe/lathe-verify/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/devenjarvis/lathe/lathe-verify"><img src="https://agentmods.dev/badge/skills/devenjarvis/lathe/lathe-verify.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Rogue Agent · line 24 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 | $0.00073 | $0.00818 |
| Opus 5 | $0.00036 | $0.00409 |
| Sonnet 5 | $0.00015 | $0.00164 |
| Haiku 4.5 | $0.00007 | $0.00082 |
Grade A, and why
lathe-verify scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lathe — Verify a Tutorial
Follow a stored tutorial exactly as a reader would, in a throwaway directory, and record whether it actually works. Triggered by /lathe-verify <slug>. Isolation is by instruction — a fresh mktemp -d, under the user's normal interactive permissions. No sandbox-exec, no Docker.
This skill is read-only with respect to the tutorial: never edit the parts or the metadata. The only writes are status updates through lathe verify-result.
Protocol
-
Mark it in-flight first:
lathe verify-result <slug> --status verifyingThis sets the spinner badge in the web UI. If it errors with "cannot verify while it is extending", stop — a part is mid-flight; don't verify on top of it.
-
Make a fresh scratch dir and work there:
cd "$(mktemp -d)"Everything the tutorial tells the reader to create happens here, not in the user's project. (Status is set by this skill, never by the web/CLI button — so an unclicked button can never strand a tutorial at
verifying.) -
Follow each part in order. Read
~/.lathe/tutorials/<slug>/part-NN.mdfrompart-01up. Install the prerequisites, create the files and paste the code blocks exactly as written, in order, then run the## Checkpointcommand and compare against the stated expected output.- Skip the pedagogical and provenance callouts —
> [!PREDICT],> [!RECALL], and> [!UNVERIFIED]are not verifiable steps. They prompt the reader or flag uncertainty; there's nothing to execute. - Treat the Checkpoint commands and code blocks as the executable surface.
- Skip the pedagogical and provenance callouts —
-
Record the terminal result with the matching command:
- Everything works →
lathe verify-result <slug> --status verified - A required tool isn't installed →
lathe verify-result <slug> --status skippedThis is not a failure — it's the ⚠️ Skipped badge, meaning "couldn't run it here," not "the tutorial is wrong." Use it whenever the toolchain the tutorial needs (compiler, runtime, SDK) is missing locally. - Something genuinely breaks (wrong output, code doesn't compile, a step contradicts itself) →
lathe verify-result <slug> --status failed \ --part <part-NN.md> \ --failed-step <1-indexed step number within that part> \ --error "<the error message or mismatched output>"
- Everything works →
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.
- 10d ago First seen · 50 lines · 73 tokens per session scan A bbe76f6aa56f
lathe-verify is a skill published in the GitHub repository devenjarvis/lathe (1,662 stars, last pushed 1mo ago), licensed MIT. It adds 73 tokens to every session and 818 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 2026-08-30.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
neuron-evaluation-engineer
Create and run AI evaluations with datasets, assertions, and output drivers in Neuron AI. Use this skill whenever the user mentions evaluation, testing AI systems, creating evaluators, dataset-driven testing, assertion-based validation, or wants to measure AI system performance. Also trigger for tasks involving…
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.
skill-benchmark
Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.