What is althing-poll?

A command runs a one-question poll across a small set of AI personas and summarizes the common themes.

Use it for fast directional feedback on an idea, message, product, or decision.

It gives a quick, structured reaction to one question without setting up a full research session. The responses are simulated rather than collected from real people.

althing-poll — Command by DataViking-Tech

Install

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.

agentmods

npx agentmods add commands/dataviking-tech/althing/althing-poll

Clone the repo

git clone --depth 1 https://github.com/DataViking-Tech/Althing

Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 711 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00014	$0.00711
Opus 5	$0.00007	$0.00356
Sonnet 5	$0.00003	$0.00142
Haiku 4.5	$0.00001	$0.00071

Measured 2d ago against content hash 9342e8351c64, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

althing-poll 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 2d 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.

commands/althing-poll.md · 76 lines

How it starts

The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Run a quick synthetic poll using the Althing MCP server. One question, a handful of AI personas, one synthesized writeup.

Question: $ARGUMENTS

What to do

Resolve the question. If $ARGUMENTS is empty, ask the user for the question before continuing. Do not invent one.
Pick personas. Default to the built-in zero-config persona set by omitting the personas argument. Only build a custom list when the user names a specific audience (e.g. "three enterprise SRE personas"). In that case, either:
- Call mcp__althing__list_persona_packs and get_persona_pack to pull an installed pack, or
- Hand-author 3–5 dicts with name, age, occupation, background, and personality_traits.
Keep panels small — run_quick_poll caps at 3 personas when running in sampling mode (no API key).
Run the poll. Call mcp__althing__run_quick_poll with:
- question: the user's question
- personas: omit for the default set, or pass your custom list
- synthesis: leave as true (default) — the synthesis pass is the point of this command
Do not pass model unless the user explicitly asked for a specific one. The MCP server picks a sensible default (haiku in BYOK mode, or the host's own model in sampling mode).

Report the result. Present the output in this shape:

Question: <the question>
Panel: <n> personas (<mode: BYOK model-name | sampling>)

## Themes
<bulleted list from the synthesis>

## Per-persona responses
- <persona name>: <one-line summary of their response>
- ...

## Notable divergences
<any disagreements the synthesis flagged>

Cost: <$X.XX from the response's `cost` field, or "n/a" in sampling mode>

Keep each persona summary to one line. The full transcripts are in the raw tool output — don't re-print them.

Guardrails

Synthetic panels are exploratory, not validating. End the report with a single-line caveat: "Synthetic panel — directional signal only, not a substitute for real user research."
Don't loop. If the user wants a second question, that's another /althing-poll invocation, not a follow-up inside this one. For multi-question structured studies, suggest the /focus-group skill instead.
Report errors honestly. If the tool returns an error field (missing API key, persona cap exceeded, etc.), show it verbatim and suggest the fix — don't silently fall back.

Read the full file on GitHub · 76 lines

Changes

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.

2d ago First seen · 76 lines · 14 tokens per session scan A 9342e8351c64

Subscribe to this mod's changes

althing-poll is a command published in the GitHub repository DataViking-Tech/Althing (2 stars, last pushed 23d ago), licensed MIT. It adds 14 tokens to every session and 711 once invoked, about $0.0001 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.

Data and AI Languages Research llm-evaluation mcp persona-simulation python survey-research synthetic-respondents

althing-poll

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

What to do

Guardrails

What this file has done since we first saw it

Other commands, from other repositories

pr-address

statusbar-style

run

init

install

language