layer4-semantic-evaluation

A method for checking software workflows from the user's point of view. It examines whether people can discover features, complete tasks efficiently, understand feedback, and recover from problems.

In plain words
What is it for?
Auditing user journeys, mapping goals to steps, and evaluating discovery, actions, problem-solving, and improvement workflows.
Why use it?
Code structure alone cannot show whether a workflow actually helps users reach their goals.

Agent

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 agents/terryc21/workflow-audit/layer4-semantic-evaluation
Clone the repo
git clone --depth 1 https://github.com/Terryc21/workflow-audit
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,988 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 $0.00000 $0.01988
Opus 5 $0.00000 $0.00994
Sonnet 5 $0.00000 $0.00398
Haiku 4.5 $0.00000 $0.00199

Measured 2d ago against content hash 52f44e69d604, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

layer4-semantic-evaluation 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.

skills/workflow-audit/agents/layer4-semantic-evaluation.md · 241 lines

How it starts

The opening of the file, as written. The whole thing — 241 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Layer 4: Semantic Evaluation

Note: Concrete examples in this file are drawn from a real workflow-audit run on the Stuffolio codebase. The evaluation dimensions (discovery, efficiency, feedback, recovery) and persona structure are reusable methodology; specific personas, workflows, and findings are project-specific. A full sample scan lives in examples/sample-stuffolio-scan/.

Purpose

Layer 4 evaluates workflows from the user's perspective. While Layers 1-3 analyze code structure, Layer 4 asks: "Does this workflow help the user achieve their goal?"

User Goals Analysis

Goal Categories

Category User Intent Example
Discovery "What can this app do?" Sees promotion card, explores
Action "I want to do X" Add item, track warranty
Problem-Solving "I have a problem" Damaged item, expiring warranty
Optimization "Make things better" Clean up, organize, export

Goal-to-Workflow Mapping

For each user goal, trace the optimal path:

goal: "Track warranty for my new Apple device"
user_type: "New user with Apple product"
optimal_workflow:
  steps: 3
  path:
    - "See AppleCare+ card on dashboard"
    - "Tap card, select device"
    - "Enter warranty details"
actual_workflow: "Matches optimal (after recent fix)"
rating: "Good"

Evaluation Criteria

Criterion 1: Goal Discoverability

Can users find the feature that helps them?

Rating Definition
Excellent Feature appears contextually when relevant
Good Feature is in logical location
Fair Feature exists but hidden
Poor Feature hard to find or missing

Criterion 2: Path Efficiency

How many steps to reach the goal?

Rating Steps Example
Optimal 1-2 Tap → Done
Acceptable 3-4 Tap → Select → Configure → Done
Suboptimal 5+ Navigation → Scroll → Search → Tap → Done
Broken Cannot complete

Read the full file on GitHub · 241 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. 2d ago First seen · 241 lines · 0 tokens per session scan A 52f44e69d604

Subscribe to this mod's changes

layer4-semantic-evaluation is an agent published in the GitHub repository Terryc21/workflow-audit (58 stars, last pushed 5d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 1,988 tokens. 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 2026-08-30.