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.
npx agentmods add instructions/aweussom/claude-code-quota/claude-mdgit clone --depth 1 https://github.com/aweussom/claude-code-quotaWrote 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.
[](https://agentmods.dev/instructions/aweussom/claude-code-quota/claude-md)<a href="https://agentmods.dev/instructions/aweussom/claude-code-quota/claude-md"><img src="https://agentmods.dev/badge/instructions/aweussom/claude-code-quota/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.02557 | $0.02557 |
| Opus 5 | $0.01278 | $0.01278 |
| Sonnet 5 | $0.00511 | $0.00511 |
| Haiku 4.5 | $0.00256 | $0.00256 |
Grade A, and why
claude-code-quota CLAUDE.md 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 6d 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.
How it starts
The opening of the file, as written. The whole thing — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — working on claude-code-quota
Invariants and traps. Read before changing either library.
What this is
Two standalone libraries — quota-lib.ps1 (Windows) and quota-lib.sh
(Linux/WSL2) — that read Claude's OAuth quota endpoint and expose the numbers to
a statusline script. They must behave identically and write a byte-compatible
cache. There is no daemon: everything happens inside a statusline refresh.
The core is small and should stay small. Most of the code is not capability, it is edge-case honesty: never show a figure we can prove is wrong, never corrupt a cache shared by several sessions, never hammer a failing endpoint.
Deployment reality (drives most of the design)
Several Claude Code sessions run at once — typically 3–5 on Windows plus 1–2 in WSL2. They all share one cache file per environment. Any change to cache reads, cache writes, or the fetch lock has to hold under concurrent access; assume 5 processes hit the same code path in the same second.
Windows and WSL2 do not share a cache: ~ is %USERPROFILE%\.claude\ in one
and /home/<you>/.claude/ in the other. That is intended — see TODONT.md.
Traps that have already bitten us
PowerShell: ConvertFrom-Json turns ISO timestamps into local-time DateTime
This is the nastiest one. Reading the cache with ConvertFrom-Json does not
give you the ISO string you wrote. It gives a DateTime in local time, and
casting that with [string] renders it in locale format with the UTC marker
gone. Re-parsing the result is then wrong by the machine's UTC offset.
On UTC+2 this made a bucket two hours from resetting look already reset,
which under the rollover rule blanked the percentage and displayed 5h:?⚠ on a
perfectly healthy account. It also silently skewed rollover detection in the
stale carry-forward path for as long as that code existed.
Rule: never [string]-cast a timestamp read from the cache. Route every
timestamp through _QL_DateToIso, which normalises both DateTime and string
inputs to ISO 8601 UTC. _QL_IsPast, _QL_GetTimeUntil and _QL_SameInstant
each call it internally and take an untyped value precisely so a caller cannot
reintroduce this — do not add a [string] type constraint to their parameters.
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.
- 6d ago First seen · 195 lines · 2,557 tokens per session scan A b4b472dfb7a9
claude-code-quota CLAUDE.md is an instructions file published in the GitHub repository aweussom/claude-code-quota (6 stars, last pushed 8d ago), licensed MIT. It adds 2,557 tokens to every session, about $0.0128 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.
Other instructions, from other repositories
claude-code-statusline CLAUDE.md
Instructions for fjlmcm/claude-code-statusline, covering claude.md, project overview, commands, architecture and core modules (scripts/lib/).
claude-scope CLAUDE.md
Instructions for YuriNachos/claude-scope, covering claude.md, quick commands, installation & default config, tech stack and architecture (overview).
cc-hud CLAUDE.md
Instructions for WaterTian/cc-hud, covering claude.md — cc-hud 项目指令, 项目概述, 技术栈, 项目结构 and 数据流.
claude-code-mascot-statusline AGENTS.md
Instructions for TeXmeijin/claude-code-mascot-statusline, covering agents.md, primary goal, what matters here, architecture and non-negotiable product rules.
rate-limit-monitor AGENTS.md
Instructions for meitalbensinai/rate-limit-monitor, covering install instructions for ai agents, what the script does (so you can explain it if asked), requirements to mention if they fail, uninstall and do not.
claude-hud-enhanced CLAUDE.md
Claude Code instructions for johnleeGitHub/claude-hud-enhanced, covering claude.md, project overview, build commands, test (node.js test runner — requires build first) and test (bun — runs source .ts directly, no build needed).