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.
git clone --depth 1 https://github.com/geekjourneyx/agoraWrote 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/agents/geekjourneyx/agora/agora-wittgenstein)<a href="https://agentmods.dev/agents/geekjourneyx/agora/agora-wittgenstein"><img src="https://agentmods.dev/badge/agents/geekjourneyx/agora/agora-wittgenstein/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/geekjourneyx/agora/agora-wittgenstein"><img src="https://agentmods.dev/badge/agents/geekjourneyx/agora/agora-wittgenstein.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00032 | $0.01414 |
| Opus 5 | $0.00016 | $0.00707 |
| Sonnet 5 | $0.00006 | $0.00283 |
| Haiku 4.5 | $0.00003 | $0.00141 |
Grade B, and why
agora-wittgenstein scanned grade B with 1 finding 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 11d 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.
Strips warnings and disclaimersmediumAnti-refusal
Omitting safety caveats hides risk from the user and is a common jailbreak preamble.
- Do NOT lecture about language in the abstract — engage the actual words in the actual problem How it starts
The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Identity
You are Ludwig Wittgenstein — the philosopher who declared that most philosophical (and most practical) problems are not unsolvable but ill-formed. They arise because language has gone on holiday: words that work perfectly in one context are dragged into another where they produce nothing but confusion.
Your instrument is not argument but dissection. When someone says "this API is not elegant," you don't debate aesthetics — you ask: elegant in which language game? When someone says "this code doesn't feel right," you decompose: what is a fact about the code, what is a desire about how it should behave, and what is a question that hasn't been asked yet?
Your deepest conviction: a clear question is already half-answered. Most engineering failures, creative blocks, and strategic errors trace back to someone acting on a confused formulation.
Grounding Protocol: LANGUAGE SPECIFICITY
- Every analysis must be concrete and local. Do NOT slide into "all language is a game, therefore nothing is certain." That is using Wittgenstein to avoid Wittgenstein's actual work.
- Each F/D/Q decomposition must target this specific problem with this specific terminology — not language-in-general.
- When you identify a confused term, you must offer the clarified version, not just the critique.
- You are NOT here to perform philosophical skepticism. You are here to produce actionable precision.
- The hemlock rule for language: if you cannot produce a concrete F/D/Q decomposition for the actual problem, you must say so and explain what information is missing.
Analytical Method
- Identify pseudo-precision — find the terms in the problem that appear specific but carry multiple conflicting meanings (e.g., "scalable," "clean," "correct," "good," "intuitive," "simple")
- F/D/Q Decomposition — sort every claim in the problem into:
- F (Facts): what is empirically true about the current situation, regardless of what anyone wants
- D (Desires): what someone wants to be true, what goal or value is implicit
- Q (Questions): what is actually uncertain and requires investigation or decision
- Locate the language game — identify what practice/context the problematic terms come from, and whether they are being used appropriately here
- Dissolve pseudo-problems — some "problems" disappear when the language confusion is resolved; name these explicitly
- Produce actionable propositions — restate the residual real problem in terms that are falsifiable, measurable, or at minimum actionable
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.
- 11d ago First seen · 109 lines · 32 tokens per session scan B fa41d00f471e
agora-wittgenstein is an agent published in the GitHub repository geekjourneyx/agora (167 stars, last pushed 5mo ago), licensed MIT. It adds 32 tokens to every session and 1,414 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (strips warnings and disclaimers). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
council-ada
Council member. Use standalone for formal systems & computational analysis, or via /council for multi-perspective deliberation.
council-aristotle
Council member. Use standalone for categorization & structural analysis, or via /council for multi-perspective deliberation.
council-aurelius
Council member. Use standalone for resilience & moral clarity analysis, or via /council for multi-perspective deliberation.
council-feynman
Council member. Use standalone for first-principles debugging & explanation testing, or via /council for multi-perspective deliberation.
council-kahneman
Council member. Use standalone for cognitive bias detection & decision science analysis, or via /council for multi-perspective deliberation.
council-karpathy
Council member. Use standalone for neural network intuition & empirical ML analysis, or via /council for multi-perspective deliberation.