New: 97,000+ 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

Evolve

by DanMcInerney in DanMcInerney/orchflows

Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable search.

MITAuto-check passedGame Development

Install Evolve

skills CLI
$ npx skills add DanMcInerney/orchflows --skill evolve -a claude-code

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

GitHub CLI
$ gh skill install DanMcInerney/orchflows evolve --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/DanMcInerney/orchflows.git skills-src && mkdir -p .claude/skills && cp -r skills-src/example-workflows/evolve/skills/evolve .claude/skills/evolve && 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
evolve
GitHub stars
117
Token cost
~805 tokens
SKILL.md length
371 words
Files
5 (incl. references)
Skills in repo
12
Repo updated
First seen
Licence
MIT

At a glance

Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable search.

  • Works in 4 steps: Start and choose. Record each… → Make. Save the plan and isolated… → Compare. Freeze candidates. Apply… → …
  • Game Development work in your project
  • SKILL.md covers Start or resume, Repeat and Return or yield
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Evolve is an agent skill from DanMcInerney/orchflows. Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable search.

Its SKILL.md is about 810 tokens, which your agent loads only when the skill is triggered. The skill folder holds 6 other files, including reference files (for example `agents/openai.yaml`, `references/evaluation.md` and `references/harness.md`).

It sits in Game Development. The repository describes itself as: 2 skills, composable into workflows, which can then be built into more complex workflows. The licence is MIT.

When your agent uses it

  • Game Development work in your project

Example prompts

  • “/evolve”

Workflow steps

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

  1. Start and choose. Record each artifact/harness attempt under core's iteration bounds and the state contract before proposal or production…
  2. Make. Save the plan and isolated candidate locations. Supply each challenger its parent snapshot, hypothesis, active harness, public…
  3. Compare. Freeze candidates. Apply screening, matched evaluation revisions, independent orchflows:orch-review judgments and confirmation…
  4. Record and continue. Journal decision/evidence, then checkpoint winner, alternatives, lessons and remaining bounds. Check stopping…

What it can do on your machine

Read from SKILL.md and the folder at commit 8d16eb3. 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

    No scripts in the folder and no shell commands in SKILL.md.

    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

Evolve loads about 805 tokens when it runs, and up to ~3.1k if it reads all its reference files. Until then it costs about 43 tokens; SKILL.md has 371 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~43
When it runs · the whole SKILL.md, loaded when a task matches
~805
With references · SKILL.md plus every file in references/, read only if the agent opens them
~3.1k

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 DanMcInerney/orchflows at commit 8d16eb3, republished under its MIT licence (© DanMcInerney). 371 words, ~805 tokens.

Download SKILL.mdSave it as .claude/skills/evolve/SKILL.md (or your agent's skills folder). This skill also uses 4 other files; get the full folder from GitHub.
name
evolve
description
Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable search.
disable-model-invocation
true

Apply library context and required evaluation and state contracts. Accept artifacts, seeds or a creation brief; purpose and constraints; optional bounds, output location and scoped settings. Infer quality criteria; clarify gaps that prevent meaningful creation or assessment. Production follows core staffing rules and evaluation's context exclusions.

Start or resume

Use the caller's directory or a distinct evolve-runs/<run-id>/ in their workspace. Reconcile saved state and in-flight work before new work.

For a new run, preserve the original or create a seed from the brief. Design, calibrate and save evaluation before challengers. Evaluate seeds, choose the incumbent (current best), and save the harness: working instructions, tools and search policy.

Record bounds, scoped settings and per-experiment limits. Default to three rounds and one challenger per round. Explicit continuous requests remove the total round cap: continue until stopped, a supplied bound is reached or execution is blocked. A target stops work only if the caller says so; plateau is not completion.

Repeat

  1. Start and choose. Record each artifact/harness attempt under core's iteration bounds and the state contract before proposal or production. Read checkpoint/evidence; choose a hypothesis and predicted benefit. When informative rounds stop yielding promotions, change approach, investigate the repeated failure or test a harness change.
  2. Make. Save the plan and isolated candidate locations. Supply each challenger its parent snapshot, hypothesis, active harness, public evaluation and guidance. Give it one owner; exclude incumbent, evaluation and journal from production writes. A harness experiment replaces the ordinary round.
  3. Compare. Freeze candidates. Apply screening, matched evaluation revisions, independent orchflows:orch-review judgments and confirmation. Compare every proposed replacement with the incumbent. Missing evidence, failed requirements, ties or unconfirmed wins retain the incumbent. Record failures without invented scores.
  4. Record and continue. Journal decision/evidence, then checkpoint winner, alternatives, lessons and remaining bounds. Check stopping conditions before another round. Repair uninformative, inconsistent or exploitable evaluation between rounds under the evaluation contract.
Show full SKILL.md (58 more words)Show less

Return or yield

Persist at each decision and before interruption/context rollover. Return best artifact, improvements/evaluation revision, active harness, evidence, observable resource use, actual stop reason and resume path. Separate measured, judge-preferred and untested claims. Missing execution/continuation capability requires a checkpoint and gap, not a background-execution claim. Reaching a finite bound completes the search without proving further improvement impossible.

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

Files

SKILL.md and 4 other files (references) in example-workflows/evolve/skills/evolve of DanMcInerney/orchflows.

  • SKILL.md
  • agents/openai.yaml
  • references/evaluation.md
  • references/harness.md
  • references/state.md

Open the folder on GitHubat commit 8d16eb3

Compare with similar skills

Evolve 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.

Evolve compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Evolve this skillDanMcInerney/orchflows117—~805Automated safety check: PassMIT
Image to Three.js Modelimg2threejs/img2threejs18k1 repos~8.2kAutomated safety check: PassApache-2.0
Web CloneJane-xiaoer/claude-skill-web-clone1k1 repos~2.7kAutomated safety check: PassMIT
Threejs Game Directormajidmanzarpour/threejs-game-skills2.4k—~2.2kAutomated safety check: PassMIT
Game Asset Generatorhtdt/godogen7.1k—~2.8kAutomated safety check: PassMIT
Threejs Gameplay Systemsvalkor-ai/loom1.2k1 repos~1.4kAutomated safety check: PassApache-2.0

Similar skills

  • Image to Three.js Model

    img2threejs/img2threejs

    Rebuilds the object in a reference image as a procedural, animation-ready Three.js model written entirely in code, using staged sculpting with quality checks.

    18k GitHub starsUsed in 1 repo~8.2k tokens
    Game DevelopmentAuto-check passed

  • Web Clone

    Jane-xiaoer/claude-skill-web-clone

    网站复刻 / 克隆方法论。USE WHEN 用户说 复刻网站、克隆网站、clone website、抄个站、仿站、 照着这个站做一个、reproduce site、还原某个网页效果、把这个站搬下来改成我的、 复刻某个交互/WebGL/Canvas/Three.js 效果。提供「先拿真源码 → 判路径 → 逆向拆解 → 搭工程 → 替换内容」的可移植决策树，覆盖静态站 /…

    1k GitHub starsUsed in 1 repo~2.7k tokens
    Game DevelopmentAuto-check passed

  • Threejs Game Director

    majidmanzarpour/threejs-game-skills

    Entrypoint for building, upgrading, and finishing Three.js browser games.

    2.4k GitHub stars~2.2k tokensUpdated 9 days ago
    Game DevelopmentAuto-check passed

  • Game Asset Generator

    htdt/godogen

    Generates game art from text prompts: PNG images, GLB 3D models, rigged characters, animations and sprites, with background removal.

    7.1k GitHub stars~2.8k tokensUpdated 6 days ago
    Game DevelopmentAuto-check passed

  • Threejs Gameplay Systems

    valkor-ai/loom

    Build and iterate playable Three.js game systems: starter scaffold, architecture, design briefs, core loops, level and encounter design, entities, input, camera, collision and physics, scoring…

    1.2k GitHub starsUsed in 1 repo~1.4k tokens
    Game DevelopmentAuto-check passed

  • Uloop Execute Dynamic Code

    CyberAgentGameEntertainment/NovaShader

    Execute C with Unity APIs when existing uloop tools cannot inspect or edit enough.

    1.6k GitHub starsUsed in 1 repo~1.9k tokens
    Game DevelopmentAuto-check passed

More from DanMcInerney/orchflows

All 12 skills in this repo

  • Benchmaker

    DanMcInerney/orchflows

    Build benchmarks of realistic, hard agent tasks, researched from real work, admitting only verified, shortcut-resistant, calibrated tasks and reporting how well the suite separates systems.

    117 GitHub stars~2k tokensUpdated 3 days ago
    Auto-check passed

  • 3D Browser Game

    DanMcInerney/orchflows

    Build a complete Three.js game or bounded production phase through mechanics experiments, Blender assets, QA and independent playtests.

    117 GitHub stars~1.5k tokensUpdated 3 days ago
    Auto-check passed

  • Design Loop

    DanMcInerney/orchflows

    Develop a project endgoal through N bounded cycles of brainstorm and research, design, implementation, comparison testing and analysis.

    117 GitHub stars~870 tokensUpdated 3 days ago
    Auto-check passed

  • Export Workflow

    DanMcInerney/orchflows

    Export an Orchflows workflow as a standalone native skill with a portability report and bounded trial.

    117 GitHub stars~674 tokensUpdated 3 days ago
    Auto-check passed

  • Gauntlet Loop

    DanMcInerney/orchflows

    Drive an ambitious goal past a real-world quality bar by splitting it into pieces, each looped through a builder and a fresh, harsh, blind critic until ours reaches the bar or the caller stops.

    117 GitHub stars~1.2k tokensUpdated 3 days ago
    Auto-check passed

  • Orchflows Review

    DanMcInerney/orchflows

    Review Orchflows against its purpose and design principles (architecture, workflow design, wording and bugs) and apply reviewed improvements on unmerged branches.

    117 GitHub stars~1k tokensUpdated 3 days ago
    Auto-check passed

Categories

Questions about Evolve

What does Evolve do?

Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable search. Evolve is an agent skill from DanMcInerney/orchflows. Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable search.

When should I use Evolve?

Evolve fits situations like: game Development work in your project.

How do I install Evolve in Claude Code?

Run `npx skills add DanMcInerney/orchflows --skill evolve -a claude-code`. Or copy the skill folder (example-workflows/evolve/skills/evolve in DanMcInerney/orchflows) into .claude/skills/evolve in your project. Claude Code loads it when a task matches its description.

How do I install Evolve in Codex?

Run `npx skills add DanMcInerney/orchflows --skill evolve -a codex`. Or copy the skill folder (example-workflows/evolve/skills/evolve in DanMcInerney/orchflows) into .agents/skills/evolve in your project. Codex loads it when a task matches its description.

Can I use Evolve 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 DanMcInerney/orchflows --skill evolve -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/evolve, .gemini/skills/evolve, .github/skills/evolve and .opencode/skills/evolve in your project.

What does Evolve need to run?

SKILL.md names no scripts, command-line tools or credentials: Evolve is instructions for the agent only.

Does Evolve 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 Evolve 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 Evolve use?

Evolve 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 Evolve use?

About 805 tokens (SKILL.md is roughly 3.2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2.3k tokens, read only when the agent opens those files.

What are the alternatives to Evolve?

Skills that share tags, products or a category with Evolve: Image to Three.js Model (img2threejs/img2threejs, 18k stars), Web Clone (Jane-xiaoer/claude-skill-web-clone, 1k stars), Threejs Game Director (majidmanzarpour/threejs-game-skills, 2.4k stars) and Game Asset Generator (htdt/godogen, 7.1k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Evolve?

DanMcInerney (a GitHub user) maintains it in DanMcInerney/orchflows, which has 117 GitHub stars. The repository holds 12 skills in this directory. The repository was last updated on October 4, 2026.

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