answerability-audit

Evaluates if a machine could construct a correct answer to an important user question from the available page evidence without guessing.

Skill by mahinigam

IndexedMIT

Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/mahinigam-answerability-audit

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

About

Evaluates if a machine could construct a correct answer to an important user question from the available page evidence without guessing.

Use cases: security

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools: No tools found
Secrets: None asked for
Install scripts: None found
Runs as: Python package

Network hosts it talks to 1

codeload.github.com · found in install command

Files it touches

~/.claude/skills/answerability-audit

What's inside

Skill name: answerability-audit
Description: Evaluates if a machine could construct a correct answer to an important user question from the available page evidence without guessing.
License: MIT
Instructions: 19 lines in SKILL.md · view on GitHub

Before installing

Ships executable scripts

Agent Skills spec issues

allowed-tools: must be a space-separated string

Files (2)

SKILL.md
audit.py · scripts/

SKILL.md (first lines)

# Answerability Audit

## When to use
Use this skill to determine if the page actually answers the user's implicit questions when referred by an AI. A page might have content, but if it lacks specific answers to basic queries (like price, features), the AI won't cite it.

## Inputs
- `AuditContext` containing sampled pages and their text content.

## Procedure
1. Define 8 synthetic question intents: `purchase_decision`, `identity`, `specification`, `trust`, `location`, `freshness`, `use_case`, and `comparison`.
2. Map triggers (URL paths and visible text) to dynamically determine relevance per page.
3. Apply a weighted scoring algorithm: critical intents (e.g., purchase with `weight=3.0`) impact the score more heavily than secondary intents (`weight=1.0`), preventing sparse but correctly focused pages from being unfairly penalized.
4. Extract facts purely from visible text (to simulate AI extraction) and score facts as 1.0 (found unambiguously), 0.5 (ambiguous), or 0.0 (missing).
5. Generate findings if the final weighted answerability score falls to or below 0.5.

## Output
Appends Findings to the `AuditContext`.

From mahinigam/skill-marketplace, MIT.

Signals

Stars: 0
Forks: 0
Open issues: 0
Last commit (this folder): Sep 9, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License: MIT
Repository: github.com/mahinigam/skill-marketplace
First seen: Oct 24, 2026
Last checked: Oct 24, 2026
Source: GitHub github-graphql

More from mahinigam

Cite this page

answerability-audit skill. Agent Mart. https://agentmart.dev/skills/mahinigam-answerability-audit (last updated 2026-10-24)