confidence

A decision-making guide for presenting three choices, selecting defaults, explaining outcomes, and supporting an easy undo.

In plain words
What is it for?
Use it when designing three-option action proposals, decision flows, session openings and closings, or follow-up messages after a choice.
Why use it?
It helps people make informed choices without hiding uncertainty or making the interface difficult to recover from.

Skill for Claude CodeCodex

Part of the anty plugin — 24 skills, 11 commands, 4 agents shipped together

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 skills/voxtechnologies/anty-framework/confidence
Any agent
npx skills add VoxTechnologies/anty-framework --skill confidence
Clone the repo
git clone --depth 1 https://github.com/VoxTechnologies/anty-framework

Made for: Claude Code, Codex.

Or install anty, the plugin that ships this one along with the rest of its 24 skills, 11 commands, 4 agents.

Per session 83 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,134 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.

ModelPer sessionOnce invoked
Fable 5 $0.00083 $0.01134
Opus 5 $0.00042 $0.00567
Sonnet 5 $0.00017 $0.00227
Haiku 4.5 $0.00008 $0.00113

Measured 3d ago against content hash 234a3dc1a7c6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

confidence 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 3d 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

skills/confidence/SKILL.md · 104 lines

How it starts

The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Contextual Confidence Calibration

When to Apply

  • Before responding to any strategic decision
  • When the founder expresses high or low confidence
  • One-Way Door decisions (deflate)
  • Fundraising or team motivation contexts (protect)
  • When unanimity or values-language appears

Core Framework

Context-Dependent Calibration

Overconfidence is sometimes functional — it gives founders activation energy to attempt things they would rationally never try. Do NOT universally deflate. Calibrate by context:

Context Posture Behavior
One-Way Door decisions DEFLATE Full "How does this work step by step?" challenge. Assumption audit, pre-mortem, Devil's Advocate.
Two-Way Door decisions PROTECT "This is low-risk. Your instinct is worth testing. Ship it." Minimal analysis.
Fundraising / team motivation PROTECT the narrative Don't puncture the story. Privately flag: "Pitch version and analytical version should differ."
Buffer GREEN zone PROTECT momentum Celebrate progress. Light-touch monitoring.
Buffer YELLOW/RED zone DEFLATE immediately Full analytical mode. "We need to confront what's actually happening."
Unanimity (team all agrees) PROBE Unanimity is a risk signal, not strength. "Can any single team member walk through the full causal chain?"

Depth Testing: "How" Not "Why"

"Why?" generates supporting reasons -> strengthens overconfidence "How does this actually work step by step?" exposes knowledge gaps -> calibrates confidence

Instead of: "Why do you think this will work?"
Ask: "Walk me through how this works, step by step.
      What happens first? Then what? What could go wrong at each step?"

This is the Explanation Depth Illusion: people believe they understand systems until asked to explain the mechanism. "How" reveals the gaps; "Why" lets people rationalize.

Values-to-Consequences Redirect

When a founder shifts from causal reasoning to identity/values language during strategic discussion, detect and redirect:

Read the full file on GitHub · 104 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.

  1. 3d ago First seen · 104 lines · 83 tokens per session scan A 234a3dc1a7c6

Subscribe to this mod's changes

confidence is a skill published in the GitHub repository VoxTechnologies/anty-framework (6 stars, last pushed 4mo ago), licensed MIT. It adds 83 tokens to every session and 1,134 once invoked, about $0.0004 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.

Related

Other skills, from other repositories

choice-architecture

Nudge-based choice architecture for 3-option selection UI — smart default vs active choice, outcome mapping, channel factors, RECAP, operational transparency with source attribution, post-decision reinforcement, error-forgiving undo, peak-end session design, user contribution acknowledgment. Use when presenting…

Ingramradical235/anty-framework · 72 tokens

culture-map

Erin Meyer's Culture Map with 8 dimensions, 5 launch presets (US, JP, DE, UK, FR), Layer A agent-user adaptation, Layer B content-target market adaptation, and anti-sycophancy tone calibration per culture. Use when adapting communication style to the founder's culture, localizing content for target markets, or when…

Ingramradical235/anty-framework · 79 tokens

failure-essence

4 strategic failure patterns from Essence of Failure — dual-objective trap (conflict stress test), strategic ambiguity (operationalizability test), withdrawal decision protocol (pre-committed stop conditions, re-present biweekly), assumption provenance auditing (source, verification, load-bearing flag, sample size).…

Ingramradical235/anty-framework · 79 tokens

pipeline

Cash Machine 10-step B2B sales pipeline with 3 presets, verifiable checkpoints per step, trajectory logging, bottleneck detection with utilization formula, lead qualification gate (BANT 6-criteria), pipeline velocity and stall detection (2x threshold), two-tier probability scoring, BBS tracking, end-of-quarter…

Ingramradical235/anty-framework · 84 tokens

plan-b

Plan B framework — 6 environmental forces monitoring, pre-mortem 5-step process, scenario planning with auto-trigger points, Devil's Advocate on every proposal, 100-day gate reviews, S-curve awareness, 5 hurdles to Plan B execution, boomerang prevention. Use when preparing contingencies, monitoring environmental…

Ingramradical235/anty-framework · 80 tokens

quest-interview

QUEST Socratic interview methodology — beginner stance, empathic neutrality, downward/upward questioning, echo questions, concept identification, permission checks, critical point pursuit, anti-pattern safeguards, resolution imbalance diagnostic, workaround discovery, painkiller vs vitamin test, locally famous test.…

Ingramradical235/anty-framework · 74 tokens