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/ymd38/dev-skills/yds-setupnpx skills add ymd38/dev-skills --skill yds-setupgit clone --depth 1 https://github.com/ymd38/dev-skillsWhat 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 | $0.00161 | $0.02682 |
| Opus 5 | $0.00081 | $0.01341 |
| Sonnet 5 | $0.00032 | $0.00536 |
| Haiku 4.5 | $0.00016 | $0.00268 |
Grade C, and why
yds-setup scanned grade C with 2 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 3d 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- **Q-H3. Bash guard strength**: standard (rm -rf /, force-push to the Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -fsSL https://raw.githubusercontent.com/ymd38/dev-skills/main/harness/scripts/install.sh \ How it starts
The opening of the file, as written. The whole thing — 210 lines — stays where its author put it; the contents beside it link to each section on GitHub.
dev-skills Full Harness Setup (Interview Mode)
Goal
Turn the current repository into a project that runs the Diagnose → Register/Draft → Plan → Resolve ⇄ Verify cycle with mechanical guardrails (hooks) and a lean CLAUDE.md.
Skills alone are L1. This setup adds L2–L3: a short project brain (CLAUDE.md), hard rails (hooks + settings), and cycle guidance (rules). It never creates scheduled loops or CI automation (L4) — that is out of scope unless the user explicitly asks in a separate task.
Principles
- Interview, no auto-detection. Do NOT infer languages from
go.mod,package.json, etc., even when they exist. This setup targets pre-development projects where detection misleads. File existence is used only for overwrite/merge prompts. - Never overwrite existing CLAUDE.md / settings / hooks without explicit per-file approval. Prefer merge (append missing sections).
- Hooks enforce hard rules; CLAUDE.md holds soft guidance. Keep
CLAUDE.md short — detail lives in skills and
.claude/rules/. - Write files only after the user approves the Phase 4 summary.
Workflow
Ask one phase at a time and wait for answers. If the user says "use defaults", still require the primary language choice, then apply the defaults table below to everything else.
How to ask (question-tool contract)
When a structured question tool (e.g. AskUserQuestion) is available, use one call per phase, and keep every call within the tool's contract — violating it aborts the turn with "Invalid tool parameters":
- At most 4 questions per call; every question needs 2–4 predefined options — never an empty or single-option list. A free-text "Other" choice is provided by the tool automatically; rely on it for custom values instead of inventing an "enter manually" option list.
- Free-form items (project name, custom commands) must still offer 2+ options: make option 1 the recommended default (e.g. "Use directory name:
- Multi-choice items (additional languages, components) use the tool's multi-select mode.
- If the tool is unavailable or returns an error, fall back to asking the same questions as plain chat text and wait for the answers — never skip a phase because the tool failed.
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.
- 3d ago First seen · 210 lines · 161 tokens per session scan C ceb5c555b78c
yds-setup is a skill published in the GitHub repository ymd38/dev-skills (3 stars, last pushed 5d ago), licensed MIT. It adds 161 tokens to every session and 2,682 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it C with 2 findings (recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
claude-md-review
Audit a CLAUDE.md file for the patterns that actually degrade Claude Code's output — vagueness, unnamed files, stale facts, and bloat. Use when asked to review, audit, improve, shrink, or fix a CLAUDE.md, and when a project's results feel inconsistent or Claude keeps rediscovering the same context.
best-practices
Searchable knowledge base of 152+ programming best practices across 30+ languages and frameworks. BM25-powered search over curated resources from industry leaders (Google, Airbnb, Uber, Mozilla, Shopify, OWASP).
common-exploit-verification
Enforce "No Exploit, No Report" policy with PoC construction standards, false-positive filtering, and evidence collection per vulnerability class across backend, frontend, and mobile. Use when validating security findings, constructing exploit proofs, filtering false positives, or writing pentest findings.
JavaScript Tooling
Development tools, linting, and testing for JavaScript projects.
common-product-requirements
Standardize PRD discovery and drafting for product scope, user outcomes, requirement IDs, and acceptance criteria. Use when creating PRD, product requirements, feature specification, or acceptance criteria plan.
system-design-building-blocks
Select infrastructure components by the constraint each removes: load balancers, caches and invalidation, queues and pub/sub, CDN, API gateway, rate limiting, consistent hashing. Use when choosing cache, queue, replica, or edge layers; defer engine tuning to the database category.