Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add eristoddle/agent-skills --skill verbalized-samplinggit clone --depth 1 https://github.com/eristoddle/agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/eristoddle/agent-skills/verbalized-sampling)<a href="https://agentmods.dev/skills/eristoddle/agent-skills/verbalized-sampling"><img src="https://agentmods.dev/badge/skills/eristoddle/agent-skills/verbalized-sampling.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00119 | $0.00833 |
| Opus 5 | $0.00060 | $0.00417 |
| Sonnet 5 | $0.00024 | $0.00167 |
| Haiku 4.5 | $0.00012 | $0.00083 |
Grade A, and why
verbalized-sampling scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verbalized Sampling
Training-free prompting technique that mitigates LLM mode collapse by generating multiple candidate responses with explicit probability assignments, then selecting from the distribution. Based on the VS research paper — achieves 1.6-2.1x diversity increase over direct prompting.
When to Apply Automatically
Apply VS internally (without the user asking) when detecting:
- Brainstorming or ideation requests
- "Give me something original/creative/unexpected"
- Tasks where the first instinct feels generic or predictable
- Requests for multiple options or alternatives
Core Workflow
1. Generate Candidates
Internally generate k candidate responses using this structure:
<response>
<text>Candidate response content.</text>
<probability>0.07</probability>
</response>
Default parameters: k=5 candidates, probability threshold tau=0.10.
2. Choose a Mode
| Mode | Instruction | Best for |
|---|---|---|
| Standard | Generate k responses with probabilities | General diversity |
| Tail | Add: "probability of each response is less than 0.10" | Maximum creativity, breaking patterns |
| CoT | Prepend: "Think step-by-step about different approaches first" | Analytical/reasoning tasks |
| Multi | Request additional distinct candidates in follow-up turns | Exhaustive exploration |
Default to Tail mode for creative tasks. Use Standard for analytical tasks. See references/prompt-templates.md for exact prompt text.
3. Select Output
After generating candidates internally, select using one of these strategies:
- Present all options — Show the user all k candidates when they want to choose (human-in-the-loop)
- Weighted random — Pick one, weighted by probabilities (default for creative tasks)
- Best-of-k — Pick highest probability (default for factual/analytical tasks)
- Tail pick — Pick lowest probability for maximum surprise
- Merge — Synthesize best elements from multiple candidates
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 81 lines · 119 tokens per session scan A 70d14de6eba3
verbalized-sampling is a skill published in the GitHub repository eristoddle/agent-skills (2 stars, last pushed 13d ago), licensed MIT. It adds 119 tokens to every session and 833 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
llm-app-patterns
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
prompt-optimization
Improve a prompt on the evaluations workbench through a measured loop. Score the baseline first, then duplicate the target column, form a hypothesis from failing rows, edit the copy's prompt draft, run, compare pass rate and cost, and repeat until the numbers hold. Use when the user asks to optimize or improve a…
enhance-prompt
Transforms vague UI ideas into polished, Stitch-optimized prompts. Enhances specificity, adds UI/UX keywords, injects design system context, and structures output for better generation results.
prompt-engineer
Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot…
seedance-vocab-en
This skill should be used when an English Seedance 2.0 prompt needs clearer production wording, less generic prose, or precise vocabulary for camera, lighting, motion, VFX, audio, and constraints. Route blocked prompts through seedance-filter for context and boundary review.
ideogram4
Prompting patterns for Ideogram 4 text-to-image — best-in-class in-image text rendering and exact color/layout control via structured JSON captions. Use when generating images that need legible on-image text (title cards, thumbnails, logos, signage, CTAs), precise brand colors, or controlled spatial layout. Triggers…