Sample several candidate answers in parallel, judge them against stated criteria, and present the winner with the runners-up kept one click away.

Apache-2.0Auto-check passed

Install Best Of N

skills CLI
$ npx skills add Sidiora-Labs/centra-gideon-agent --skill best-of-n -a claude-code

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

GitHub CLI
$ gh skill install Sidiora-Labs/centra-gideon-agent best-of-n --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/Sidiora-Labs/centra-gideon-agent.git skills-src && mkdir -p .claude/skills && cp -r skills-src/runtime/gideon/extensions/skills/bundled/best-of-n .claude/skills/best-of-n && 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
best-of-n
GitHub stars
217
Token cost
~1.2k tokens
SKILL.md length
620 words
Files
1
Skills in repo
24
Repo updated
First seen
Licence
Apache-2.0

At a glance

Sample several candidate answers in parallel, judge them against stated criteria, and present the winner with the runners-up kept one click away.

  • The user wants options rather than one answer
  • SKILL.md covers Activation Behavior, The confirmation gate (never…, Running it and Presenting the result, plus 2 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Include give me N options

What it does

Best Of N is an agent skill from Sidiora-Labs/centra-gideon-agent. Sample several candidate answers in parallel, judge them against stated criteria, and present the winner with the runners-up kept one click away. Confirms the count and the criteria first because each candidate is a separate model call. Use when the user wants options rather than one answer. Triggers include "give me N options", "give me 3 versions", "best of", "try a few versions", "sample and pick", "generate a few and choose", "which of these is best", "draft some variations".

Its SKILL.md is about 1.2k 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 companion AI agent that learns, adapts and gets the work done no matter the task. The licence is Apache-2.0.

When your agent uses it

  • The user wants options rather than one answer
  • Include give me N options
  • Give me 3 versions
  • Try a few versions

Example prompts

  • “give me N options”
  • “give me 3 versions”
  • “best of”
  • “/best-of-n”

What it can do on your machine

Read from SKILL.md and the folder at commit ca531ee. 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 (its code samples are markdown).

    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

Best Of N loads about 1.2k tokens when it runs. Until then it costs about 124 tokens; SKILL.md has 620 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~124
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

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 Sidiora-Labs/centra-gideon-agent at commit ca531ee, republished under its Apache-2.0 licence (© Sidiora-Labs). 620 words, ~1,216 tokens.

Download SKILL.mdSave it as .claude/skills/best-of-n/SKILL.md (or your agent's skills folder).
name
best-of-n
description
Sample several candidate answers in parallel, judge them against stated criteria, and present the winner with the runners-up kept one click away. Confirms the count and the criteria first because each candidate is a separate model call. Use when the user wants options rather than one answer. Triggers include "give me N options", "give me 3 versions", "best of", "try a few versions", "sample and pick", "generate a few and choose", "which of these is best", "draft some variations".
triggers
best of, give me options, give me versions, versions and pick, few versions, sample and pick, generate a few, pick the best, which is best, draft variations…

Activation Behavior

Explicit triggers (the user has already asked for several candidates):

  • "give me 3 versions and pick the best", "best of 5", "sample and pick", "generate a few options and choose"
  • Activate immediately, then run the confirmation gate below for the two things you cannot guess: the count and the criteria.

Ambiguous triggers (the user might want one good answer, not a slate):

  • "try a few versions", "draft some variations", "which of these is best", "give me options"
  • Offer the choice BEFORE spending anything:

    Want me to sample a few candidates and judge them (each candidate is its own model call), or just write you one answer?

  • If they pick one answer: answer normally. No skill mode, no sampling.

The confirmation gate (never skip the cost line)

Best-of-N costs N model calls plus one judge pass per surviving candidate. That is the whole trade, so say it out loud and get both inputs:

Sampling 3 candidates — that's 3 model calls (plus judging), not one. Judging on: specific, under 60 characters, no hype. Want a different count (max 5) or different criteria?

  • N is capped at 5. If the user asks for more, say the cap and sample 5.
  • Criteria are what "best" means for this task. If the user has not said, propose criteria in the gate rather than inventing them silently — a slate judged on an unstated bar is a coin flip with extra steps.
  • One round trip is enough. Do not interrogate; propose sensible defaults and let the user correct them.

Running it

Call the best_of_n tool once with prompt (the ask, identical for every candidate), n, and criteria. The tool fans the N calls out in parallel — one slate costs roughly one call's wall time, N calls' spend — and returns JSON:

{"winner": "...", "winner_idx": 1, "candidates": [{"idx", "temperature", "text", "error"}],
 "judgments": [{"idx", "score", "reason"}], "judged": true, "n": 3, "note": ""}

Do NOT re-sample to "double-check", and do not run it twice for the same ask. The slate you already have is the slate.

Show full SKILL.md (302 more words)Show less

Presenting the result

Lead with the winner as a normal answer — the user asked for a good answer, not a report. Then keep the slate one click away:

markdown
<winner text, presented as the answer>

<details>
<summary>Runners-up (2) — say "use #2" to switch</summary>

**#1** — score 3.5 · *reason from the judge*

<candidate text>

**#3** — score 4.0 · *reason from the judge*

<candidate text>

</details>

Numbering rules, because "use #2" has to actually work:

  • Number candidates #1…#N in slate order: #k is the candidate whose idx is k - 1. Never renumber by score — the user's "#2" must mean the same candidate in your list and in the tool result.
  • Mark the winner in the visible answer (e.g. "picked #2 of 3"), so the numbers the user sees are the numbers they can choose from.
  • When the user says "use #2" (or "go with the second one"), reply with that candidate's text verbatim from the slate you already have. Do not re-sample, do not paraphrase, do not re-judge. If they then ask you to edit it, edit that text.

Degraded cases — report them, never paper over them

  • judged: false — the judge was unavailable. Say so: "the judge wasn't available, so this is the first candidate, unranked" and still show the slate.
  • Some candidates carry an error — the slate is narrower than N. Say how many returned ("2 of 3 came back") and judge only those. Never invent a missing candidate.
  • winner: null — every sample failed. Say that plainly and offer to answer directly or retry. Do not present one of the errors as an answer.

When NOT to use this

  • Anything with one correct answer (a fact, a calculation, a file's contents). N samples of a lookup is N× the cost for the same answer.
  • Long-running or tool-using work — this is one-shot text sampling, not delegation. Use the delegation skill and subagents for work that needs tools.
  • Inside a loop over many items. One slate per user ask; a slate per item multiplies the spend by N without anyone noticing.

© Sidiora-Labs, Apache-2.0. 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 runtime/gideon/extensions/skills/bundled/best-of-n of Sidiora-Labs/centra-gideon-agent.

Open the folder on GitHubat commit ca531ee

Compare with similar skills

Best Of N 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.

Best Of N compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Best Of N this skillSidiora-Labs/centra-gideon-agent217—~1.2kAutomated safety check: PassApache-2.0
Parallels Discord Roundtripopenclaw/openclaw392k—~788Automated safety check: PassMIT
Openclaw Parallels Smokeopenclaw/openclaw392k—~8.4kAutomated safety check: NotesMIT
Parallel Execution Optimizeraffaan-m/ECC277k1 repos~712Automated safety check: PassMIT
Parallel AutomationComposioHQ/awesome-claude-skills77k3 repos~734Automated safety check: PassNone
Best-of-N Candidate Tournamentcodewhale-hq/Codewhale41k—~1.2kAutomated safety check: PassMIT

Similar skills

  • Run macOS Parallels smoke with Discord send, host verification, host reply, and guest readback proof.

    392k GitHub stars~788 tokensUpdated today
    Auto-check passed
  • Openclaw Parallels Smoke

    openclaw/openclaw

    Prepare, snapshot, run, rerun, debug, or interpret OpenClaw Parallels guest install, onboarding, gateway smoke, and upgrade checks across macOS, Windows, and Linux.

    392k GitHub stars~8.4k tokensUpdated today
    Auto-check: notes
  • Speed up a task by turning it into a dependency graph of parallel lanes with a lane matrix, batched reads and checks, write surfaces isolated by file, worktree, branch, or service, and a final…

    277k GitHub starsUsed in 1 repo~712 tokens
    DevelopmentAuto-check passed
  • Parallel Automation

    ComposioHQ/awesome-claude-skills

    Automate Parallel tasks via Rube MCP (Composio). An agent skill from ComposioHQ/awesome-claude-skills.

    77k GitHub starsUsed in 3 repos~734 tokens
    Productivity & AutomationAuto-check passed
  • Best-of-N Candidate Tournament

    codewhale-hq/Codewhale

    Generates several independent candidate solutions in parallel worktrees, judges them once against one explicit rubric, and applies the winning candidate only after it passes verification.

    41k GitHub stars~1.2k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Parallel Web

    K-Dense-AI/scientific-agent-skills

    Uses Parallel CLI for web search, URL extraction, deep research, structured data enrichment, entity discovery, and recurring web monitoring.

    48k GitHub starsUsed in 1 repo~2.2k tokens
    Research & ScienceAuto-check: notes

More from Sidiora-Labs/centra-gideon-agent

All 24 skills in this repo
  • Artifacts

    Sidiora-Labs/centra-gideon-agent

    Persist, version, and iterate on LLM-generated UI (widgets, HTML, markdown).

    217 GitHub stars~1.2k tokensUpdated 3 days ago
    Auto-check passed
  • Document Authoring

    Sidiora-Labs/centra-gideon-agent

    Author a real document — a report, proposal, slide deck, spec, or paper — by settling audience, claim, and structure before drafting prose, then writing to that structure.

    217 GitHub stars~970 tokensUpdated 3 days ago
    Auto-check passed
  • Editorial Document

    Sidiora-Labs/centra-gideon-agent

    Write long-form editorial documents as clean, semantic HTML saved as an artifact (kind=document) — reports, briefs, explainers, proposals, write-ups.

    217 GitHub stars~1.3k tokensUpdated 3 days ago
    Auto-check passed
  • Gideon API

    Sidiora-Labs/centra-gideon-agent

    The operator's manual for DRIVING Gideon's API and tools — how to find an exact tool/route signature, the mandatory verify-after-mutate loop, and what NOT to hand-roll.

    217 GitHub stars~1.4k tokensUpdated 3 days ago
    Auto-check passed
  • Infographic Syntax

    Sidiora-Labs/centra-gideon-agent

    Author infographics with the AntV declarative DSL — pick a template, fill a small indented data tree, and save as an artifact (kind=infographic) that renders to crisp SVG and streams as you write it.

    217 GitHub stars~1.1k tokensUpdated 3 days ago
    Auto-check passed
  • Research Campaign

    Sidiora-Labs/centra-gideon-agent

    Run a multi-cycle research campaign — grill the question, decompose it into answerable sub-questions, investigate each against sources, then synthesize with confidence and gaps stated.

    217 GitHub stars~1.3k tokensUpdated 3 days ago
    Auto-check passed

Questions about Best Of N

What does Best Of N do?

Sample several candidate answers in parallel, judge them against stated criteria, and present the winner with the runners-up kept one click away. Best Of N is an agent skill from Sidiora-Labs/centra-gideon-agent. Sample several candidate answers in parallel, judge them against stated criteria, and present the winner with the runners-up kept one click away.

When should I use Best Of N?

Best Of N fits situations like: the user wants options rather than one answer; include give me N options; give me 3 versions; try a few versions.

How do I install Best Of N in Claude Code?

Run `npx skills add Sidiora-Labs/centra-gideon-agent --skill best-of-n -a claude-code`. Or copy the skill folder (runtime/gideon/extensions/skills/bundled/best-of-n in Sidiora-Labs/centra-gideon-agent) into .claude/skills/best-of-n in your project. Claude Code loads it when a task matches its description.

How do I install Best Of N in Codex?

Run `npx skills add Sidiora-Labs/centra-gideon-agent --skill best-of-n -a codex`. Or copy the skill folder (runtime/gideon/extensions/skills/bundled/best-of-n in Sidiora-Labs/centra-gideon-agent) into .agents/skills/best-of-n in your project. Codex loads it when a task matches its description.

Can I use Best Of N 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 Sidiora-Labs/centra-gideon-agent --skill best-of-n -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/best-of-n, .gemini/skills/best-of-n, .github/skills/best-of-n and .opencode/skills/best-of-n in your project.

What does Best Of N need to run?

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

Does Best Of N 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 Best Of N 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 Best Of N use?

Best Of N is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Best Of N use?

About 1.2k tokens (SKILL.md is roughly 4.9k 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 Best Of N?

Skills that share tags, products or a category with Best Of N: Parallels Discord Roundtrip (openclaw/openclaw, 392k stars), Openclaw Parallels Smoke (openclaw/openclaw, 392k stars), Parallel Execution Optimizer (affaan-m/ECC, 277k stars) and Parallel Automation (ComposioHQ/awesome-claude-skills, 77k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Best Of N?

Sidiora-Labs (a GitHub organization) maintains it in Sidiora-Labs/centra-gideon-agent, which has 217 GitHub stars. The repository holds 24 skills in this directory. The repository was last updated on October 8, 2026.

Source: Sidiora-Labs/centra-gideon-agent on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.