What is cache-architecture?

A specification for quota-axi's local cache: where quota snapshots are stored, how permissions are set, how accounts and environments are separated, and how records are retained. A cache is saved data used to avoid fetching the same information repeatedly.

Use it when implementing or reviewing quota caching, cache paths, file permissions, provider context identifiers, cleanup, and protection of sensitive data.

It prevents cached data from crossing account boundaries and prevents secrets such as tokens, keys, cookies, or raw responses from being stored. Strict file permissions also limit local access.

cache-architecture — Skill for Claude Code by kunchenguid/quota-axi

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 kunchenguid/quota-axi --skill cache-architecture

Clone the repo

git clone --depth 1 https://github.com/kunchenguid/quota-axi

Made for: Claude Code.

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/kunchenguid/quota-axi/cache-architecture/github.svg)](https://agentmods.dev/skills/kunchenguid/quota-axi/cache-architecture)

Your own site

<a href="https://agentmods.dev/skills/kunchenguid/quota-axi/cache-architecture"><img src="https://agentmods.dev/badge/skills/kunchenguid/quota-axi/cache-architecture/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/kunchenguid/quota-axi/cache-architecture"><img src="https://agentmods.dev/badge/skills/kunchenguid/quota-axi/cache-architecture.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 25 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,136 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	$0.00025	$0.01136
Opus 5.5	$0.00010	$0.00454
Sonnet 5.5	$0.00005	$0.00227
Haiku 4.5	$0.00003	$0.00114

Measured 13d ago against content hash 3a406a18e3c0, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

cache-architecture 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 13d 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/cache-architecture/SKILL.md · 62 lines

How it starts

The opening of the file, as written. The whole thing — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Quota Cache Architecture & Context Scoping

This document details the disk caching architecture, file permissions, context-scoping mechanism, and data retention policies implemented in src/cache.ts.

Cache Storage & Security

Path: ~/.cache/quota-axi/quotas.json, or $XDG_CACHE_HOME/quota-axi/quotas.json if XDG_CACHE_HOME is set.
File Permissions: Cache files and parent directories must be created with strict 0600 (read/write by owner only) and 0700 permissions.
Normalized Data Only: Cache records contain only normalized, non-secret quota snapshot structures.
Strict Prohibition: Never persist raw HTTP responses, headers, authorization bearers, refresh tokens, API keys, or browser cookies.

Context-Scoped Providers (`CONTEXT_SCOPED_PROVIDERS`)

To prevent cross-account or cross-environment cache poisoning, providers whose identity is configuration-dependent or slot-shared are enrolled in CONTEXT_SCOPED_PROVIDERS in src/cache.ts:

Enrolled Providers: Claude, Kimi, Command Code, MiniMax, ElevenLabs, Devin, Muse, and Codex.
Context Identifiers: Snapshots carry an opaque SHA-256 context identifier derived from the answering configuration:
- Claude: Hashed profile and credential-storage selection. claudeCredentialContextId appends an env marker so environment-token readings cannot be served as stale cache for stored-token accounts.
- Kimi: Home path plus the slot and base URL resolved from config.toml (never the file contents, which contain keys). Prevents mainland China snapshots from serving as global fallbacks.
- Command Code: Winning source plus the account identity validated by whoami.
- MiniMax: Answering credential source plus deployment host.
- ElevenLabs: Credential source plus a one-way SHA-256 digest of the answering API key.
- Devin: Answering source, the first-party host it was sent to, and a one-way SHA-256 digest of the answering API key; the key never enters the cache.
- Muse: Answering source plus a one-way SHA-256 digest of the credential that answered (the Muse CLI's stored OAuth access token or an exported META_API_KEY); a since-refreshed token is a cache miss, never a cross-attribution between accounts.
- Codex: Hashed ChatGPT account ID stored by the specific credential that produced the reading (resolving collisions between lone discovered lanes and Pi entries).
Fallback Verification: Stale cache fallback rejects legacy unstamped snapshots or snapshots whose context ID does not match the active configuration.
Codex exception: Codex stamps are optional at write time, but stale serving requires ownership. readCachedCodexProvider withholds an unstamped snapshot, and withholds a snapshot whose stored account id is not among the ids the failed reading's tried credentials actually named; a sibling never probed does not vouch for it, and a transient failure is vouched for only by its own credential. A reading whose tried credentials name no account establishes nothing and is withheld from stale fallback. Contract: README Cache.
Skip on Unconfirmed: If a reading cannot confirm context identity (e.g. unreadable configuration), quota-axi skips both writing to cache and clearing the cache slot.

Read the full file on GitHub · 62 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.

13d ago First seen · 62 lines · 25 tokens per session scan A 3a406a18e3c0

Subscribe to this mod's changes

cache-architecture is a skill published in the GitHub repository kunchenguid/quota-axi (155 stars, last pushed yesterday), licensed MIT. It adds 25 tokens to every session and 1,136 once invoked, about $0.0001 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 2026-10-05.

Databases Security

create-pr

Creates a GitHub PR with a Linear-ticket-prefixed title and a decision-led, narrative description for Prisma 8. Use when the user wants to create a pull request, open a PR, or submit changes for review.

prisma/orm · 48 tokens

schema-exploration

Lists tables, describes columns and data types, identifies foreign key relationships, and maps entity relationships in a database. Use when the user asks about database schema, table structure, column types, what tables exist, ERD, foreign keys, or how entities relate.

langchain-ai/deepagents · 57 tokens

atlas-mssql

SQL Server specifics for Atlas (also Azure SQL and MSSQL): login, dev database and URL scope (mode=schema vs mode=database, the database query parameter), compatibility level and collation, system-generated constraint names, clustered index defaults and datetime2 precision in HCL, GO batches and the delimiter…

ariga/atlas · 147 tokens

ha-data-stores

Map of Hope Agent's local data stores and safe read-only query workflow. Use when the user asks where Hope Agent stores data, wants to inspect sessions/messages/memory/logs/background jobs/knowledge indexes/settings, asks the model to query local app data, or debugging requires checking persisted state. Trigger…

shiwenwen/hope-agent · 115 tokens

nornicdb-cypher-queries

Pick fast, predictable Cypher query shapes in NornicDB — point lookups, batch retrieval, pagination, search, traversal, batched UNWIND/MERGE writes, cleanup, multi-tenant isolation. Use when writing or reviewing Cypher whose latency or throughput matters; maps user intent to the executor's hot-path query templates.

orneryd/NornicDB · 79 tokens

volcengine-rds-postgresql

A tool for operating PostgreSQL databases hosted by Volcano Engine's managed database service. PostgreSQL is a relational database used to store structured application data.

bytedance/agentkit-samples · 63 tokens