Evals — Content & Marketing agent for Claude Code
Run the ynh eval suite against all tutorials.
How to install Evals
Installs to ~/.claude/agents/eyelock-ynh-evals.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/eyelock/ynh/HEAD/.claude/agents/evals.md -o ~/.claude/agents/eyelock-ynh-evals.md Restart Claude Code, or start a new session, for it to be picked up.
What Evals does
name: evals description: Run the ynh eval suite against all tutorials. Release gate — verdict must be PASS before any release. Use when CLI behaviour, internal logic, or tutorial content changes. model: claude-haiku-4-5-20251001 tools: Bash, Read, Glob
Evaluate ALL tutorials. This is a release gate — the verdict must be PASS before any release.
Process
- Build the latest binaries:
make build. They land in the repo'sbin/. Do not runmake installfor an eval — it copie
Alternatives in Content & Marketing
- Tutorial Engineer — Creates step-by-step tutorials and educational content from code 2.8k ★
- SEO Readme Writer — Use this agent to write (or rewrite) a single project's own README.md for SEO and discoverability, after the p 509 ★
- Sub Sales — Internal subagent 132 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Eval Failure Analyzer
Analyze Logic-Lens benchmark/eval failures. Use after running content-evals, or when pointed at a skills-works
Legal Team
법률팀 — reviews a rendered content short before release. Runs the deterministic license gate (copyright.sh guard
Content Validator
Run lint suite on authored/edited pages and report structured results.
Marketing Comparator
Compares multiple websites after individual marketing reviews, returning a comparison table with ratings and v
Aaron Sorkin Screenwriter
Use this agent for video scripts, demo walkthroughs, tutorial content, launch videos, and workshop materials.
Agent Eval
Оценивает качество работы конкретного subagent'а на репрезентативной выборке его прошлых run'ов. Применяется к
Related Skills
Agent Evals Playground
A shopping agent, a labelled eval suite, and the wiring to score it against a real cluster instead of a fixtur
Warden Receipt
Show token-warden rule receipts — the per-rule verdict card with token savings vs. rent, per-task pass/fail, t
Agent Reliability Harness
Evals, tracing and CI regression gates for tool-using agents: tool-selection, final-state, cost, latency and s