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

Harness Learn

by ruvnet in ruvnet/ruflo

Run a GEPA learning cycle via metaharness learn (upstream ADR-235, metaharness@0.3.0) — optimizes a harness genome against a SWE-bench-style slice manifest.

MITAuto-check: notesResearch & Science

Install Harness Learn

skills CLI
$ npx skills add ruvnet/ruflo --skill harness-learn -a claude-code

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

GitHub CLI
$ gh skill install ruvnet/ruflo harness-learn --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/ruvnet/ruflo.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/ruflo-metaharness/skills/harness-learn .claude/skills/harness-learn && 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
harness-learn
GitHub stars
74k
Used in
1 other repo
Token cost
~800 tokens
SKILL.md length
337 words
Files
1
Skills in repo
264
Repo updated
First seen
Licence
MIT

At a glance

Run a GEPA learning cycle via metaharness learn (upstream ADR-235, metaharness@0.3.0) — optimizes a harness genome against a SWE-bench-style slice manifest.

  • Works in 5 steps: Validate --repo exists when given;… → Invoke the pinned metaharness binary… → Default timeouts: 120s dry-run, 600s… → …
  • Tasks that involve Bioinformatics
  • SKILL.md covers When to use, Preconditions (upstream design), Algorithm and Cost note, plus 1 more section
  • Calls git and node; reaches github.com

What it does

Harness Learn is an agent skill from ruvnet/ruflo. Run a GEPA learning cycle via metaharness learn (upstream ADR-235, metaharness@0.3.0) — optimizes a harness genome against a SWE-bench-style slice manifest. $0 dry-run by default; --run is the explicit spend opt-in. Requires a metaharness repo checkout (--repo or $METAHARNESSREPO) — without one it reports checkout-required with clone instructions. Degrades gracefully when metaharness is absent.

Its SKILL.md is about 800 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Research & Science, covering Bioinformatics and Architecture decision records. The repository describes itself as: 🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory…. The licence is MIT.

When your agent uses it

  • Tasks that involve Bioinformatics
  • Tasks that involve Architecture decision records

Example prompts

  • “/harness-learn”

Requirements

  • Docker
  • Pre-approved tools (allowed-tools): Bash

Workflow steps

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

  1. Validate --repo exists when given; export it as $METAHARNESS_REPO.
  2. Invoke the pinned metaharness binary (metaharness@~0.4.1, local install
  3. Default timeouts: 120s dry-run, 600s with --run — real runs on larger
  4. Detect the checkout-required message → structured payload, exit 0.
  5. Parse the trailing JSON report when upstream emits one; otherwise return

What it can do on your machine

Read from SKILL.md and the folder at commit 6051f67. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Bash

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • node

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

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • github.com

    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

Harness Learn loads about 800 tokens when it runs. Until then it costs about 105 tokens; SKILL.md has 337 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~105
When it runs · the whole SKILL.md, loaded when a task matches
~800

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: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
    allowed-tools: Bash

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 ruvnet/ruflo at commit 6051f67, republished under its MIT licence (© ruvnet). 337 words, ~800 tokens.

Download SKILL.mdSave it as .claude/skills/harness-learn/SKILL.md (or your agent's skills folder).
name
harness-learn
description
Run a GEPA learning cycle via `metaharness learn` (upstream ADR-235, metaharness@0.3.0) — optimizes a harness genome against a SWE-bench-style slice manifest. $0 dry-run by default; `--run` is the explicit spend opt-in. Requires a metaharness repo checkout (`--repo` or $METAHARNESS_REPO) — without one it reports `checkout-required` with clone instructions. Degrades gracefully when metaharness is absent.
allowed-tools
Bash
argument-hint
--host <h> --model <m> --slice <manifest> [--repo <checkout>] [--run] [--alert-on-fail]

Surfaces metaharness learn — the upstream GEPA learning harness that evolves harness policy genomes against a scored task corpus instead of hand-editing prompts. Candidates are scored on held-out slices and only measured winners promote (the shipped cand-6 genome is the first such promotion: holdout gold 2/12 → 3/12, zero regressions).

When to use

  • A harness's policy prompt underperforms on a task family and you want a measured improvement loop rather than manual prompt iteration.
  • Pricing a learning run before committing spend — the default dry-run resolves the slice manifest and reports cost without any model calls.
  • After a learn run promotes a genome: pair with harness-gepa --op render to inspect what the promoted policy actually says.

Preconditions (upstream design)

The learning harness (GEPA + SWE-bench + Docker) is too heavy for the npm package, so learn needs a local clone:

bash
git clone https://github.com/ruvnet/metaharness.git
node scripts/learn.mjs --repo ./metaharness --host claude-code --model haiku --slice slices/lite.json

Without a checkout the script emits {status: "checkout-required"} and exits 0 — a precondition report, not an error (distinct from degraded: true, which means the npm package itself is absent). The managed-service path (gateway-side learn jobs, no checkout) is upstream's ADR-235 follow-up and not available yet.

Algorithm

Implementation: scripts/learn.mjs.

  1. Validate --repo exists when given; export it as $METAHARNESS_REPO.
  2. Invoke the pinned metaharness binary (metaharness@~0.4.1, local install or one-time versioned cache — never @latest): metaharness learn --host <h> --model <m> --slice <s> [--run] via _harness.mjs (graceful degradation, hard timeout).
  3. Default timeouts: 120s dry-run, 600s with --run — real runs on larger slices need an explicit --timeout-ms matched to slice size × model cost.
  4. Detect the checkout-required message → structured payload, exit 0.
  5. Parse the trailing JSON report when upstream emits one; otherwise return the raw report text under rawReport.

Cost note

--run is the ONLY path that spends. Everything else — dry-run, checkout probe, degraded path — is $0. The MCP tool (metaharness_learn) has a 120s subprocess budget; run real learning cycles from a terminal via ruflo metaharness learn ... --run --timeout-ms <big>.

Exit codes

  • 0 — report produced (or dry-run, checkout-required, degraded)
  • 1 — --alert-on-fail and the learn run reported failure
  • 2 — config error (bad --repo path)

© ruvnet, 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 plugins/ruflo-metaharness/skills/harness-learn of ruvnet/ruflo.

Open the folder on GitHubat commit 6051f67

Used in 1 other repository

We found 1 copy of this SKILL.md (exact, near-identical or edited) in other folders, from 1 other GitHub owner. This page covers the copy in ruvnet/ruflo, which our catalogue first saw on October 7, 2026.

Compare with similar skills

Harness Learn 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.

Harness Learn compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Harness Learn this skillruvnet/ruflo74k1 repos~800Automated safety check: NotesMIT
Repo Genomeruvnet/metaharness690—~772Automated safety check: PassMIT
Bio Seq ObjectsGPTomics/bioSkills1.2k1 repos~2.5kAutomated safety check: PassMIT
Chi Reproducibilitybrycewang-stanford/Awesome-Journal-Skills1.2k—~1.4kAutomated safety check: PassMIT
Dbsnp Databasegoogle-deepmind/science-skills3.2k3 repos~3.4kAutomated safety check: NotesApache-2.0
Metabolic Study Planneraiming-lab/AutoResearchClaw15k—~1.9kAutomated safety check: PassMIT

Similar skills

  • Repo Genome

    ruvnet/metaharness

    7-section readiness scorecard for a LOCAL repo. An agent skill from ruvnet/metaharness.

    690 GitHub stars~772 tokensUpdated yesterday
    Product & Project ManagementAuto-check passed

  • Bio Seq Objects

    GPTomics/bioSkills

    Create and manipulate Seq, MutableSeq, and SeqRecord objects using Biopython.

    1.2k GitHub starsUsed in 1 repo~2.5k tokens
    Research & ScienceAuto-check passed

  • Chi Reproducibility

    brycewang-stanford/Awesome-Journal-Skills

    A skill your agent uses when strengthening research transparency for an ACM CHI paper — protocols, instruments, codebooks, analysis scripts, preregistration, and data availability under…

    1.2k GitHub stars~1.4k tokensUpdated 11 days ago
    Research & ScienceAuto-check passed

  • Dbsnp Database

    google-deepmind/science-skills

    A skill your agent uses when you want to look up, map, and search for short genetic variants (SNPs, indels) in NCBI's dbSNP database.

    3.2k GitHub starsUsed in 3 repos~3.4k tokens
    Research & ScienceAuto-check: notes

  • Metabolic Study Planner

    aiming-lab/AutoResearchClaw

    Turns a broad metabolic modelling topic into a concrete, paper-shaped plan with organism, model, perturbations, metrics and figures before any FBA code is written.

    15k GitHub stars~1.9k tokensUpdated 1 mo ago
    Research & ScienceAuto-check passed

  • 13C Metabolic Flux Analysis

    K-Dense-AI/scientific-agent-skills

    Estimates reaction fluxes inside cells from steady-state carbon-13 labeling data with a bundled mfapy-based solver, and reports which fluxes the data pin down.

    48k GitHub starsUsed in 1 repo~3.2k tokens
    Research & ScienceAuto-check passed

More from ruvnet/ruflo

All 264 skills in this repo

  • AgentDB Pattern Memory

    ruvnet/ruflo

    Stores, searches, and retrieves successful patterns with HNSW-indexed semantic search so agents can reuse past solutions instead of relearning them.

    74k GitHub starsUsed in 3 repos~830 tokens
    Auto-check passed

  • Security Audit Scanner

    ruvnet/ruflo

    Runs claude-flow CLI security scans for input validation, path traversal, SQL injection, XSS, hardcoded secrets and known CVEs, and writes an audit report.

    74k GitHub starsUsed in 2 repos~823 tokens
    Auto-check passed

  • SPARC Development Methodology

    ruvnet/ruflo

    Applies the SPARC method (specification, pseudocode, architecture, refinement, completion) with 17 specialized modes and multi-agent orchestration, from research to deployment.

    74k GitHub starsUsed in 2 repos~829 tokens
    Auto-check passed

  • Swarm Orchestration

    ruvnet/ruflo

    Coordinates a hierarchical swarm of specialized agents through the claude-flow CLI for work that spans several files or modules at once.

    74k GitHub starsUsed in 2 repos~779 tokens
    Auto-check passed

  • Ruflo Multi-Agent Orchestration

    ruvnet/ruflo

    Sets up and drives Ruflo, an npm-installed orchestration layer for multi-agent swarms, persistent memory, routing, hooks and its MCP tool catalog.

    74k GitHub starsUsed in 1 repo~975 tokens
    Auto-check passed

  • Agent Coordination

    ruvnet/ruflo

    Reference for spawning, listing, monitoring and stopping agents with claude-flow commands, with agent type families, routing codes and coordination tips.

    74k GitHub starsUsed in 2 repos~519 tokens
    Auto-check passed

Categories

Questions about Harness Learn

What does Harness Learn do?

Run a GEPA learning cycle via metaharness learn (upstream ADR-235, metaharness@0.3.0) — optimizes a harness genome against a SWE-bench-style slice manifest. Harness Learn is an agent skill from ruvnet/ruflo.0) — optimizes a harness genome against a SWE-bench-style slice manifest.

When should I use Harness Learn?

Harness Learn fits situations like: tasks that involve Bioinformatics; tasks that involve Architecture decision records.

How do I install Harness Learn in Claude Code?

Run `npx skills add ruvnet/ruflo --skill harness-learn -a claude-code`. Or copy the skill folder (plugins/ruflo-metaharness/skills/harness-learn in ruvnet/ruflo) into .claude/skills/harness-learn in your project. Claude Code loads it when a task matches its description.

How do I install Harness Learn in Codex?

Run `npx skills add ruvnet/ruflo --skill harness-learn -a codex`. Or copy the skill folder (plugins/ruflo-metaharness/skills/harness-learn in ruvnet/ruflo) into .agents/skills/harness-learn in your project. Codex loads it when a task matches its description.

Can I use Harness Learn 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 ruvnet/ruflo --skill harness-learn -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/harness-learn, .gemini/skills/harness-learn, .github/skills/harness-learn and .opencode/skills/harness-learn in your project.

What does Harness Learn need to run?

Going by SKILL.md and its folder, Harness Learn needs the command-line tools its instructions call (git and node). Our summary lists: Docker. Its frontmatter pre-approves these tools: Bash.

Does Harness Learn access the network?

SKILL.md names 1 domain. In commands or code: github.com; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.

Is Harness Learn safe to install?

Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Harness Learn use?

Harness Learn 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 Harness Learn use?

About 800 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.

What are the alternatives to Harness Learn?

Skills that share tags, products or a category with Harness Learn: Repo Genome (ruvnet/metaharness, 690 stars), Bio Seq Objects (GPTomics/bioSkills, 1.2k stars), Chi Reproducibility (brycewang-stanford/Awesome-Journal-Skills, 1.2k stars) and Dbsnp Database (google-deepmind/science-skills, 3.2k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Harness Learn?

ruvnet (a GitHub user) maintains it in ruvnet/ruflo, which has 74,089 GitHub stars. The repository holds 264 skills in this directory. The repository was last updated on October 8, 2026.

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