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/klimentij/klimkit/klimkit-walkthroughnpx skills add klimentij/klimkit --skill klimkit-walkthroughgit clone --depth 1 https://github.com/klimentij/klimkitWrote 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/klimentij/klimkit/klimkit-walkthrough)<a href="https://agentmods.dev/skills/klimentij/klimkit/klimkit-walkthrough"><img src="https://agentmods.dev/badge/skills/klimentij/klimkit/klimkit-walkthrough.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 | $0.00057 | $0.00574 |
| Opus 5 | $0.00028 | $0.00287 |
| Sonnet 5 | $0.00011 | $0.00115 |
| Haiku 4.5 | $0.00006 | $0.00057 |
Grade A, and why
klimkit-walkthrough 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 — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Klimkit Walkthrough
Create a small static HTML report that lets a human inspect what happened without reading the whole conversation.
Workflow
- Ensure Klimkit's docs-first context exists. If the repo lacks the
docs/work/layout, useklimkit-setupfirst. - Define the walkthrough purpose: feature proof, bug reproduction, QA pass, setup proof, or decision brief.
- Write the report as a self-contained single-file HTML, numbered like any other artifact in the current
docs/work/phase folder (e.g.003-walkthrough.html). Keep heavy media in the phase folder's gitignored.local/subfolder unless it is irreproducible evidence. - Capture fresh evidence. Use screenshots for UI states; add video only when interaction timing or motion matters.
- Write a concise report with:
- title and timestamp;
- source task, PR, or request links;
- environment and URL under test;
- ordered steps;
- expected versus observed result;
- embedded screenshots/video as full-width sections;
- redactions and privacy notes;
- validation summary and residual risk.
- Validate that image/video links resolve from the report path.
- Deploy the report for the human driver through the harness's native hosted surface by default:
- Claude Code: publish the HTML with the built-in Artifact tool (authenticated claude.ai page, private to the author by default; republish to the same URL on revisions).
- Codex: deploy with Codex Sites (workspace-authenticated, private by default; save a version, then deploy).
- Always keep the default authenticated visibility (private or workspace-only); never enable public link sharing unless the user explicitly asks.
- If no native hosted surface is available in the session, fall back to
klimkit-report-serverfor local or Tailscale serving. - Add the report path and the deployed (or served) URL to the phase
LOG.mdand the final handoff.
Redaction Rules
- Do not expose secrets, tokens, private chat, customer data, private repo names, or tailnet internals in public artifacts.
- If sensitive material is necessary evidence, keep it local/ignored and summarize safely in tracked HTML.
- Do not claim a report is served until the served path is checked.
What ships with it
2 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 First seen · 40 lines · 57 tokens per session scan A c1561c991ebb
klimkit-walkthrough is a skill published in the GitHub repository klimentij/klimkit (7 stars, last pushed 26d ago), licensed MIT. It adds 57 tokens to every session and 574 once invoked, about $0.0003 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
production-prep
Última etapa. Revisão técnica (ortografia, gramática, consistência) + formatação para ebook e impresso. Dois especialistas num só. Usar após todas as revisões de conteúdo estarem prontas, antes de publicar ou submeter.
book-auto
One-command book generation. Give it an idea, get a publish-ready book. Dispatches the book-orchestrator agent which runs the entire 17-phase pipeline autonomously, pausing only at 3 human checkpoints.
expense-report-generation
How to turn raw card and bank statement downloads into a defensible, client-ready expense report. Use whenever the user asks to build an expense report, reconcile AmEx or bank statements, work out "what is this charge for", aggregate reimbursables for a client or trip, split a claim by month, or render an expense PDF…
lossless-flavored-markdown
The Lossless Group's extended-markdown flavor — what LFM is, when to use it, how its directives normalize across syntaxes (CommonMark, GFM, Obsidian callouts, remark-directive, Markdoc), how citations and link previews work, and how sites extend it via trigger maps and theme tokens. Use whenever authoring or rendering…
tmcp-repo-behavior-spec-loop
Use TMCP for repo-wide behavior specification loops: code-derived feature inventory, canonical spreadsheet tracking, running-app verification, defect remediation, regression coverage, and final evidence audits.
html-output
Convert the current plan, spec, design doc, or conversation context into a single self-contained, share-able HTML file with rich visual structure (tables, SVG diagrams, code blocks, responsive layout, in-page navigation). Supports four output shapes — editorial, technical-doc, dashboard, and slide-deck. Use when the…