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 human-avatar/skills-for-humanity --skill s4h-probability-confidence-calibrationgit clone --depth 1 https://github.com/human-avatar/skills-for-humanityWrote 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/human-avatar/skills-for-humanity/s4h-probability-confidence-calibration)<a href="https://agentmods.dev/skills/human-avatar/skills-for-humanity/s4h-probability-confidence-calibration"><img src="https://agentmods.dev/badge/skills/human-avatar/skills-for-humanity/s4h-probability-confidence-calibration/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/skills/human-avatar/skills-for-humanity/s4h-probability-confidence-calibration"><img src="https://agentmods.dev/badge/skills/human-avatar/skills-for-humanity/s4h-probability-confidence-calibration.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 30 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00064 | $0.01077 |
| Opus 5 | $0.00032 | $0.00539 |
| Sonnet 5 | $0.00013 | $0.00215 |
| Haiku 4.5 | $0.00006 | $0.00108 |
Grade A, and why
s4h-probability-confidence-calibration 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 9d 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 — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Probability Confidence Calibration
Overconfidence is the most documented and costly bias in judgment. People who say they are 90% confident are right far less than 90% of the time. But underconfidence is also costly — excessive hedging prevents commitment and action when evidence is actually sufficient. Calibration is not about being less confident; it is about having confidence levels that match the available evidence.
Your Process
Step 1: State the Claim and Current Confidence Name the specific claim — not a vague domain but a specific, falsifiable statement — and state the current confidence level as a percentage.
Framing check: Confirm the specific claim and confidence level before continuing. State what you've identified — the actual claim being evaluated and its stated confidence percentage — in one sentence, then use AskUserQuestion:
- Question: "I'm reading this as: [your one-sentence framing of the specific claim and confidence level, e.g. 'You believe X with 85% confidence and want to know if that's calibrated to the evidence']. Is that right?"
- Header: "Framing"
- Options:
- Yes — proceed — framing is correct
- Adjust — one element is off; user will correct it before you continue
- Reframe — different situation than read; incorporate the correction before proceeding
Step 2: Audit Supporting Evidence List the evidence supporting the claim. For each piece: classify its strength:
- Direct observation: you or a trusted source directly witnessed this
- Inference: logical or empirical inference from other data
- Anecdote: one or a few cases, not systematic
- Assumption: believed without verification
Step 3: List Counter-Evidence and Gaps What evidence exists against the claim? What would you expect to see if the claim were true that you do not see? What have you not checked that bears on the claim?
Step 4: Identify the Most Likely Failure Mode If this claim is wrong, what is the most probable reason? Is that failure mode being appropriately weighted in the current confidence assessment, or is it being minimized?
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.
- 9d ago First seen · 97 lines · 64 tokens per session scan A 9ef2425ee003
s4h-probability-confidence-calibration is a skill published in the GitHub repository human-avatar/skills-for-humanity (223 stars, last pushed 1mo ago), licensed MIT. It adds 64 tokens to every session and 1,077 once invoked, about $0.0003 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 2026-09-03.
Other skills, from other repositories
decision-making-under-pressure
Provides practical habits for making better decisions faster -- reframing problems before solving them, knowing when to stop deliberating, balancing intuition with analysis, resisting 'just this once' compromises, and breaking overthinking cycles. Use when facing a high-stakes choice with incomplete data, sensing that…
flow-state
Optimizes individual and team productivity using Csikszentmihalyi's flow state framework, covering the conditions for flow (clear goals, immediate feedback, challenge-skill balance), the four-stage flow cycle, and organizational flow design. Use when auditing personal or team focus time, diagnosing productivity loss…
stoic-right-action
Applies Stoic ethics of right action to entrepreneurial decision-making, covering the four cardinal virtues (courage, justice, temperance, wisdom), ethical leadership under pressure, and the discipline of choosing the harder right over the easier wrong. Use when facing ethical dilemmas in business, building a…
bakeoff
Turn one decision into a judged tournament of solutions, then pick the best. Given a problem, design choice, or a suggestion you want cross-verified, it generates diverse candidate solutions, auto-derives the evaluation dimensions for THAT problem (so you don't have to know what to score on), judges every candidate…
balanced-scorecard
Translate strategy into metrics across Financial, Customer, Internal Process, and Learning & Growth perspectives.
customer-journey-map
Visualize every touchpoint a customer has with a product or service — stages, actions, emotions, pain points, opportunities.