What is arch-latency-cost?

A guide for reducing response time, model calls, token use, and cost in Kore.ai Agent Platform agents. It uses session data and evaluation runs to identify expensive turns and measure changes.

Use it to measure calls, tokens, latency, and estimated cost; find the most expensive conversations; change one part of the agent; and compare the result with a baseline.

It helps explain why an agent is slow or expensive without making changes that reduce answer quality. An evaluation set is a fixed group of test conversations used to check whether changes still work.

arch-latency-cost — Skill for Codex by Koredotcom/agents-mcp-tools

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 Koredotcom/agents-mcp-tools --skill arch-latency-cost

Clone the repo

git clone --depth 1 https://github.com/Koredotcom/agents-mcp-tools

Made for: 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/koredotcom/agents-mcp-tools/arch-latency-cost/github.svg)](https://agentmods.dev/skills/koredotcom/agents-mcp-tools/arch-latency-cost)

Your own site

<a href="https://agentmods.dev/skills/koredotcom/agents-mcp-tools/arch-latency-cost"><img src="https://agentmods.dev/badge/skills/koredotcom/agents-mcp-tools/arch-latency-cost/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/koredotcom/agents-mcp-tools/arch-latency-cost"><img src="https://agentmods.dev/badge/skills/koredotcom/agents-mcp-tools/arch-latency-cost.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 91 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,560 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.

Model	Per session	Once invoked
Fable 5.1	$0.00091	$0.01560
Opus 5.5	$0.00036	$0.00624
Sonnet 5.5	$0.00018	$0.00312
Haiku 4.5	$0.00009	$0.00156

Measured 1mo ago against content hash 3b681d371f3b, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

arch-latency-cost 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 1mo 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/arch-latency-cost/SKILL.md · 109 lines

How it starts

The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Latency and cost

A turn's time is roughly the number of sequential model calls times the time each takes, plus tools and checks. Count the calls before touching prompts or models. Never optimize without an eval set (arch-eval-catalog): a faster agent that fails more conversations is not an improvement.

The loop

Measure model calls, input tokens and latency per turn for the top intents, from real or eval sessions. Use debug_session_history (sortBy: "tokenCount" or "estimatedCost", range), then debug_analyze_session and debug_get_span_tree on the worst sessions (pass projectId and source: "historical" for sessions that are not live).
Rank turns by frequency × cost.
Change one lever from the table below.
Re-run the eval set and platform_eval_runs compare with the baseline: success must hold, cost must drop.
Record before and after numbers in the change note.

Where a turn's time goes

Step	Runs when	Lever
Input checks	Every turn with input guardrails	Rule-based checks first; model judges only where meaning matters
Routing call	Every turn that starts at a router or orchestrator	Deterministic `WHEN`; a faster model tier for coordination
Main model calls	1, plus 1 per tool round	Fewer tool rounds; smaller prompts
Tool calls	Each round	Fast APIs, parallel calls, `MODEL_CONTEXT`, tight timeouts
Tool-result summaries	A result over 10,000 chars under `summarize`	`MODEL_CONTEXT` or `structured`
Extraction	GATHER fields outside inline gather	`EXECUTION.inline_gather: true`
Output checks	Every reply with output guardrails	Fewer output judges; a streaming-safe mode
Reask	An output rule with reask fires	`block` with a good message on latency-critical agents
Handoff hops	Each handoff	Flatter topology; let orchestrators answer simple asks

Read the full file on GitHub · 109 lines

Files

What ships with it

2 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.

1mo ago First seen · 109 lines · 91 tokens per session scan A 3b681d371f3b

Subscribe to this mod's changes

arch-latency-cost is a skill published in the GitHub repository Koredotcom/agents-mcp-tools (4 stars, last pushed 1mo ago), licensed MIT. It adds 91 tokens to every session and 1,560 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-10-09.

Analytics and metrics Data and AI

arch-latency-cost next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
arch-latency-costKoredotcom/agents-mcp-tools	Skill	4	1mo ago	MIT	A	91	Codex
agent-platform-model-registrygoogle/skills	Skill	24k	today	Apache-2.0	A	60	Claude Code, Codex
llm-evaluationsickn33/agentic-awesome-skills	Skill	50k	today	MIT	A	25	Claude Code
foundry-config-setupmicrosoft/agent-framework	Skill	15k	today	MIT	A	65	Claude Code, Codex
google-cloud-solution-agentic-analytics-spark-knowledge-cataloggoogle/skills	Skill	24k	today	Apache-2.0	A	132	Claude Code, Codex
training-checkwanshuiyin/Auto-claude-code-research-in-sleep	Skill	19k	2d ago	MIT	A	35	Claude Code
nemo-automodel-launcher-configNVIDIA/skills	Skill	4.0k	today	Apache-2.0	A	30	Claude Code