cost-performance-scorer

cost-performance-scorer is a skill for Claude Code, Codex from AnthonyAlcaraz/agentic-graph-rag-skills. It costs 155 tokens per session (2,219 once invoked), scanned A, original, MIT.

A scoring method for comparing how different AI models balance cost and result quality when assigned different tasks.

In plain words
What is it for?
Use it to evaluate model-routing policies, compare models for specific workflow steps, and detect when a cheaper model fails too often.
Why use it?
It measures the cost of successfully completed work instead of token prices alone, and accounts for cases where some mistakes matter more than others.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to evaluate model-routing policies, compare models for specific workflow steps, and detect when a cheaper model fails too often.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer
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.

Any agent
npx skills add AnthonyAlcaraz/agentic-graph-rag-skills --skill cost-performance-scorer
Clone the repo
git clone --depth 1 https://github.com/AnthonyAlcaraz/agentic-graph-rag-skills

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for cost-performance-scorer

README.md
[![agentmods](https://agentmods.dev/badge/skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer/github.svg)](https://agentmods.dev/skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer)
Your own site
<a href="https://agentmods.dev/skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer"><img src="https://agentmods.dev/badge/skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer/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.

agentmods 80×15 button for cost-performance-scorer

Your own site · 80×15
<a href="https://agentmods.dev/skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer"><img src="https://agentmods.dev/badge/skills/anthonyalcaraz/agentic-graph-rag-skills/cost-performance-scorer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 155 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,219 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00155 $0.02219
Opus 5 $0.00077 $0.01110
Sonnet 5 $0.00031 $0.00444
Haiku 4.5 $0.00015 $0.00222

Measured 11d ago against content hash dc6348e4aa85, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

cost-performance-scorer 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 11d ago.

The scan reads SKILL.md. This mod also ships 2 executable files (cli.py, lib.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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/optimization/cost-performance-scorer/SKILL.md · 162 lines

How it starts

The opening of the file, as written. The whole thing — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Cost-Performance Scorer

Overview

Routing strategies are only as good as the data that informs them. Selective intelligence works only if you can measure it, and the book names two metrics that matter:

  • Cost per successful completion. Cost per task completed correctly, not cost per token. A cheap model that fails 40% of the time is not cheaper than an expensive model that succeeds on the first attempt — the wasted spend on failures is amortized over the successes.
  • Quality parity threshold. The minimum acceptable quality per node type. The AlertClassifier may need 0.99 (a wrong validation is worse than none); the QueryAnalyst may tolerate 0.90 because downstream nodes recover from misclassification.

The book's harness wraps every node and logs a NodeInvocation per call (Example 8-3). Success is judged per node — matching a human-labeled severity for the AlertClassifier, an SRE accepting the recommendation for PredictionSynthesis. Each candidate model is scored against a per-node evaluation set with domain failure weights (Example 8-4): a P1 alert misclassified as P3 is 10x worse than the reverse, an asymmetry a generic accuracy metric cannot express. Kakao's AI Shopping Mate is the worked anchor: restructuring GPT-4o-everywhere into a workflow graph of fine-tuned 27-32B models moved format adherence from 0.655 to 0.987 and accuracy from 0.578 to 0.890, with the smaller models preferred over GPT-4o in 63% of 2,100 turns.

When to Use

  • You have run production traffic through a routed pipeline and need to know which node is over- or under-provisioned.
  • You are choosing a specialist SLM and must evaluate it on your data, not MMLU.
  • A cheap node looks cheap per token but you suspect its failure rate erases the savings.
  • You need a per-node quality gate before promoting a routing change.

Phrases that should invoke this skill: "cost per successful completion", "is this model actually cheaper", "per-node evaluation set", "quality parity threshold", "score the routing policy", "weighted error rate".

Read the full file on GitHub · 162 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 11d ago First seen · 162 lines · 155 tokens per session scan A dc6348e4aa85

Subscribe to this mod's changes

cost-performance-scorer is a skill published in the GitHub repository AnthonyAlcaraz/agentic-graph-rag-skills (10 stars, last pushed 2mo ago), licensed MIT. It adds 155 tokens to every session and 2,219 once invoked, about $0.0008 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

graphify

Use for any question about a codebase, its architecture, file relationships, or project content — especially when graphify-out/ exists, where the question should be treated as a graphify query first. Turns any input (code, docs, papers, images, videos) into a persistent knowledge graph with god nodes, community…

Graphify-Labs/graphify · 76 tokens

lemmalog

Externalize working memory and logical state into the lemmalog Datalog engine (MCP). Use for ANY multi-step task where state should outlive one context window or span agents: long investigations, debugging sessions, audits, multi-agent searches, systematic explorations, planning with many interdependent constraints…

JordyZomer/lemmalog · 105 tokens

jurisd-research

Expert Australian/NZ legal research and AGLC4 citation using the jurisd MCP server. Use when finding cases or legislation (AustLII), looking up a provision offline, formatting or resolving citations, building a pinpoint, tracing who-cites-what, or producing an AGLC4 bibliography. Triggers on case law, legislation…

russellbrenner/jurisd · 133 tokens

repo_search

Search repository text through a deterministic first-class fak command.

anthony-chaudhary/fak · 14 tokens

container-manager-kg-ingestion

Snapshot a host's Docker/Podman/Swarm inventory into the epistemic-graph knowledge graph as typed OWL nodes via the container-manager-mcp MCP server — containers, images, volumes, networks, swarm services and nodes, with their :usesImage / :runsOn / :builtFrom links. Use when the agent must record live container state…

Knuckles-Team/container-manager-mcp · 116 tokens

cortex-design

Use this skill to generate well-branded interfaces and assets for Cortex, either for production or throwaway prototypes/mocks/etc. Contains essential design guidelines, colors, type, fonts, assets, and UI kit components for prototyping.

mocaOS/cortex-app · 50 tokens