New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Jev Blocks

by autonomous-ai in autonomous-ai/openharness

Design a falling-blocks challenge for Jev in blocks.json (gravity, speedup, garbage, a custom board, the piece mix, the style line), then read the verdict and report where Jev starts to break.

MITAuto-check passed

Install Jev Blocks

skills CLI
$ npx skills add autonomous-ai/openharness --skill jev-blocks -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install autonomous-ai/openharness jev-blocks --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/autonomous-ai/openharness.git skills-src && mkdir -p .claude/skills && cp -r skills-src/store/agents/jev-blocks/skills/blocks .claude/skills/jev-blocks && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
jev-blocks
GitHub stars
1.1k
Token cost
~1k tokens
SKILL.md length
579 words
Files
1
Skills in repo
100
Repo updated
First seen
Licence
MIT

At a glance

Design a falling-blocks challenge for Jev in blocks.json (gravity, speedup, garbage, a custom board, the piece mix, the style line), then read the verdict and report where Jev starts to break.

  • Works in 4 steps: Fix the seed and set speedup to 0. → Change one knob, save, let the game run… → Read .harness/verdict.json. Look at the… → …
  • SKILL.md covers The decision loop, The honest limit, Knobs and what they do and How to find the breaking point, plus 2 more sections
  • Calls node

What it does

Jev Blocks is an agent skill from autonomous-ai/openharness. Design a falling-blocks challenge for Jev in blocks.json (gravity, speedup, garbage, a custom board, the piece mix, the style line), then read the verdict and report where Jev starts to break.

Its SKILL.md is about 1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: The ultimate harness for coding agents and beyond. All your agents. All your machines. One command center. Start with code, then follow your curiosity and build across… The licence is MIT.

Example prompts

  • “/jev-blocks”

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. Fix the seed and set speedup to 0.
  2. Change one knob, save, let the game run for a minute or so.
  3. Read .harness/verdict.json. Look at the summary line and the timing finding
  4. Move the knob and repeat. Report the last setting that played clean and the first that did not.

What it can do on your machine

Read from SKILL.md and the folder at commit 50da5db. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • node

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Jev Blocks loads about 1k tokens when it runs. Until then it costs about 51 tokens; SKILL.md has 579 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~51
When it runs · the whole SKILL.md, loaded when a task matches
~1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from autonomous-ai/openharness at commit 50da5db, republished under its MIT licence (© autonomous-ai). 579 words, ~1,040 tokens.

Download SKILL.mdSave it as .claude/skills/jev-blocks/SKILL.md (or your agent's skills folder).
name
jev-blocks
description
Design a falling-blocks challenge for Jev in blocks.json (gravity, speedup, garbage, a custom board, the piece mix, the style line), then read the verdict and report where Jev starts to break.

Craft: a Jev Blocks challenge

Jev places every piece of a falling-blocks puzzle with one call. The craft is to set up a game that shows a real decision loop under a real limit, then to find the point where the limit wins.

The decision loop

For each new piece the viewer builds one state text:

  • the style line,
  • the well as 20 ASCII rows (# filled, . empty),
  • column heights, holes, bumpiness and the deepest well,
  • the current piece and the next three,
  • level, lines, gravity, and what a decision and a move cost,
  • a table of every legal placement. The id is r<rotation>c<leftmost column>, for example r1c4.

Then it asks three questions in the same call:

idtypewhat it asks
placechoiceone option per placement. Each option's description is a short fact line: lines cleared, holes, column heights, max height, bumpiness, wells, moves needed
dangerscore, 4 levelshow close the stack is to topping out
go_for_fournoulis it worth keeping a column open for an I piece

Questions cannot see each other. Shared facts live in the state text.

The honest limit

Time. The piece falls at gravity x (1 + speedup x (level - 1)) rows per second. A decision costs decisionMs. Each rotate or shift costs moveMs. The piece falls the whole time. When it lands, it locks, even if it has not reached the chosen column. The pane marks that as TOO SLOW and counts it under MISSED.

A rough rule: a placement needs decisionMs + (moves + 1) x moveMs. The fall gives free rows / gravity seconds. With 110 ms and 35 ms, a wall placement (about 6 moves) needs about 355 ms. On an empty well that fits up to about 46 rows per second. On a stack 8 rows high it only fits up to about 28. So height and gravity work together, and garbage makes speed bite sooner.

Show full SKILL.md (264 more words)Show less

Knobs and what they do

jsonc
{
  "gravity": 2,        // 0.2 to 40. The starting speed.
  "speedup": 0.5,      // 0 to 2. 0 keeps gravity fixed, which is best for a clean comparison.
  "decisionMs": 110,   // 30 to 1000. Raise it to play a slower model.
  "moveMs": 35,        // 5 to 500.
  "garbageRows": 0,    // 0 to 12.
  "weights": { "I": 1, "O": 1, "T": 1, "S": 1, "Z": 1, "J": 1, "L": 1 },
  "board": [],         // up to 20 rows of 10 characters, '.' or '#', top row first
  "style": "Keep the stack low and flat. Go for four when it is safe."
}

Ideas that make a good challenge:

  • Speed test. speedup: 0, then try gravity 10, 25, 40. Add garbageRows: 8 so 40 really bites.
  • Slow thinker. Keep gravity at 12 and set decisionMs to 300, then 700. Watch the misses start.
  • Dig-out. garbageRows: 10, low gravity, a style line that says dig first.
  • Cruel mix. "weights": { "S": 1, "Z": 1 }, or drop the I piece and see what go_for_four says.
  • Set piece. Draw a board that is one I piece from four lines and see if Jev takes it.

How to find the breaking point

  1. Fix the seed and set speedup to 0.
  2. Change one knob, save, let the game run for a minute or so.
  3. Read .harness/verdict.json. Look at the summary line and the timing finding (N of the last 50 pieces landed short of their target at G rows/s) and the topout finding.
  4. Move the knob and repeat. Report the last setting that played clean and the first that did not.

Checking

bash
node "$JEV_DSH/toolchain/check.mjs"

It prints ok or one error line per problem: a number out of range, weights that are all zero, a board row that is not 10 characters of . and #.

Honesty

  • The game is synthetic. The score means nothing outside this pane.
  • Without an API key the pane says MOCK. The offline reader is a fixed heuristic. It shows that the plumbing works. It does not show how well live Jev plays. Say which one you measured.
  • Never edit .harness/verdict.json, and never tell the person to open a browser or start a server.

© autonomous-ai, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in store/agents/jev-blocks/skills/blocks of autonomous-ai/openharness.

Open the folder on GitHubat commit 50da5db

Compare with similar skills

Jev Blocks next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Jev Blocks compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Jev Blocks this skillautonomous-ai/openharness1.1k—~1kAutomated safety check: PassMIT
Block Kitopenclaw/openclaw392k—~624Automated safety check: PassMIT
Agent Challengesruvnet/ruflo74k3 repos~995Automated safety check: PassMIT
Challengealirezarezvani/claude-skills28k2 repos~1.7kAutomated safety check: PassMIT
Render Blockingthedaviddias/Front-End-Checklist74k—~430Automated safety check: PassMIT
Blocking IO Guardbytedance/deer-flow83k—~1.7kAutomated safety check: PassMIT

Similar skills

  • Block Kit

    openclaw/openclaw

    Use proactively for structured or interactive Slack replies, and when asked to author or validate native Slack Block Kit JSON.

    392k GitHub stars~624 tokensUpdated today
    Auto-check passed

  • Agent Challenges

    ruvnet/ruflo

    Agent skill for challenges - invoke with $agent-challenges. An agent skill from ruvnet/ruflo.

    74k GitHub starsUsed in 3 repos~995 tokens
    Auto-check passed

  • Challenge

    alirezarezvani/claude-skills

    Pre-mortem plan analysis. An agent skill from alirezarezvani/claude-skills.

    28k GitHub starsUsed in 2 repos~1.7k tokens
    Business, Finance & HRAuto-check passed

  • Render Blocking

    thedaviddias/Front-End-Checklist

    A skill your agent uses when auditing slow page loads, heavy assets, or rendering delays related to Eliminate render-blocking resources.

    74k GitHub stars~430 tokensUpdated 2 days ago
    Frontend & DesignAuto-check passed

  • Blocking IO Guard

    bytedance/deer-flow

    Adds a runtime test anchor for backend async code that could block the asyncio event loop, and proves the anchor fails when the blocking call returns.

    83k GitHub stars~1.7k tokensUpdated yesterday
    DevelopmentAuto-check passed

  • Color Blocking

    sickn33/agentic-awesome-skills

    Web and App implementation guide for Color Blocking. An agent skill from sickn33/agentic-awesome-skills.

    47k GitHub starsUsed in 1 repo~2.4k tokens
    MobileAuto-check passed

More from autonomous-ai/openharness

All 100 skills in this repo

  • G-code Slicer Tool

    autonomous-ai/openharness

    Slices 3D mesh files into printer-profiled plain G-code through real slicer CLIs, with backend discovery, input inspection, dry runs and static validation.

    1.1k GitHub stars~1.5k tokensUpdated today
    Auto-check passed

  • Home Assistant Automation Builder

    autonomous-ai/openharness

    Turns a home-automation request into standard, testable automations.yaml, run against Home Assistant Core's real triggers and verified with its own trace tool.

    1.1k GitHub stars~1.2k tokensUpdated today
    Auto-check passed

  • Score Music Composer

    autonomous-ai/openharness

    Turns a musical brief into LilyPond concert-pitch music, checked parts for each instrument and a playable practice pack.

    1.1k GitHub stars~1.1k tokensUpdated today
    Auto-check passed

  • OrcaSlicer 3MF and G-code Workflow

    autonomous-ai/openharness

    Turns an STL and explicit printer and material requirements into compared OrcaSlicer plans, an editable 3MF project, checked G-code and a portable handoff.

    1.1k GitHub stars~1.5k tokensUpdated today
    Auto-check passed

  • Sheets and Docs Report Builder

    autonomous-ai/openharness

    Builds an editable DOCX report, a formula-driven XLSX workbook and a fresh LibreOffice PDF preview from one structured source file, then checks them together.

    1.1k GitHub stars~708 tokensUpdated today
    Auto-check passed

  • Bambu Labs

    autonomous-ai/openharness

    Dry-run, upload, and cautiously initiate local Bambu Lab print jobs from validated plain .gcode, using Bambu LAN FTPS/MQTT handoffs.

    1.1k GitHub starsUsed in 1 repo~2.1k tokens
    Auto-check: warnings

Questions about Jev Blocks

What does Jev Blocks do?

Design a falling-blocks challenge for Jev in blocks.json (gravity, speedup, garbage, a custom board, the piece mix, the style line), then read the verdict and report where Jev starts to break. Jev Blocks is an agent skill from autonomous-ai/openharness.json (gravity, speedup, garbage, a custom board, the piece mix, the style line), then read the verdict and report where Jev starts to break.

How do I install Jev Blocks in Claude Code?

Run `npx skills add autonomous-ai/openharness --skill jev-blocks -a claude-code`. Or copy the skill folder (store/agents/jev-blocks/skills/blocks in autonomous-ai/openharness) into .claude/skills/jev-blocks in your project. Claude Code loads it when a task matches its description.

How do I install Jev Blocks in Codex?

Run `npx skills add autonomous-ai/openharness --skill jev-blocks -a codex`. Or copy the skill folder (store/agents/jev-blocks/skills/blocks in autonomous-ai/openharness) into .agents/skills/jev-blocks in your project. Codex loads it when a task matches its description.

Can I use Jev Blocks in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add autonomous-ai/openharness --skill jev-blocks -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/jev-blocks, .gemini/skills/jev-blocks, .github/skills/jev-blocks and .opencode/skills/jev-blocks in your project.

What does Jev Blocks need to run?

Going by SKILL.md and its folder, Jev Blocks needs the command-line tools its instructions call (node).

Does Jev Blocks access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Jev Blocks safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Jev Blocks use?

Jev Blocks is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Jev Blocks use?

About 1k tokens (SKILL.md is roughly 4.2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Jev Blocks?

Skills that share tags, products or a category with Jev Blocks: Block Kit (openclaw/openclaw, 392k stars), Agent Challenges (ruvnet/ruflo, 74k stars), Challenge (alirezarezvani/claude-skills, 28k stars) and Render Blocking (thedaviddias/Front-End-Checklist, 74k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Jev Blocks?

autonomous-ai (a GitHub organization) maintains it in autonomous-ai/openharness, which has 1,149 GitHub stars. The repository holds 100 skills in this directory. The repository was last updated on October 8, 2026.

Source: autonomous-ai/openharness on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.