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 skills add avenoxai/avenoxskills --skill limitgit clone --depth 1 https://github.com/avenoxai/avenoxskillsWrote 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/skills/avenoxai/avenoxskills/limit)<a href="https://agentmods.dev/skills/avenoxai/avenoxskills/limit"><img src="https://agentmods.dev/badge/skills/avenoxai/avenoxskills/limit/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.
<a href="https://agentmods.dev/skills/avenoxai/avenoxskills/limit"><img src="https://agentmods.dev/badge/skills/avenoxai/avenoxskills/limit.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00050 | $0.01135 |
| Opus 5.5 | $0.00020 | $0.00454 |
| Sonnet 5.5 | $0.00010 | $0.00227 |
| Haiku 4.5 | $0.00005 | $0.00113 |
Grade A, and why
limit 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 26d 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 — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Limit
Run the script that ships next to this file (python instead of python3 on Windows;
adjust the path if you installed the skill somewhere else):
python3 ~/.claude/skills/limit/scripts/limit.py --color always
Show the output verbatim before summarising it — do not paraphrase the numbers.
A window whose warnings include stale, window expired, or live fetch failed
is unknown, not empty. Don't treat its percentage as current.
Acting on it
Each window has its own wall, and the reset time matters as much as the percentage:
| Reading | Do |
|---|---|
| Both windows of a pool clear | Use the pool normally. Don't hedge or narrate the number. |
| A window within ~5 points of full | Narrow the work: fewer parallel jobs, cheapest model that can carry it, sequential. Say the number once. |
| 5-hour window at the wall and resets within ~30 min | Wait for the reset. Don't hand off to the other pool or start a long run that will be cut in half. |
| Weekly window at the wall | It won't come back for days: hand the work to the other pool if both of its windows are clear, otherwise keep it in the main loop or queue it. |
| Both pools at a wall, or a window unknown | Delegate nothing. Report both numbers and ask. |
The fullest valid window in each pool is binding — a 12% 5-hour window next to an exhausted weekly window means the pool is exhausted. Keep a few percent in reserve for whatever must run at the end of a session (handoff notes, summaries): a job that dies mid-flight costs more than one that never started.
Sources, in order
- Claude: live
https://api.anthropic.com/api/oauth/usagecall using the token in~/.claude/.credentials.json→ optional--claude-observation PATHfile →~/.claude.jsoncache (cachedUsageUtilization, only refreshed by an interactive/usagecall, can be hours stale). - Codex: CodexBar's local cache:
%APPDATA%/CodexBar/codex-accounts/snapshots.jsonon Windows,~/Library/Application Support/CodexBar/codex-account-snapshots.jsonon macOS (Swift Codable format), or$XDG_CONFIG_HOME/CodexBar/codex-accounts/snapshots.jsonon Linux (~/.configby default, when a compatible writer exists). Then the newest dated corerate_limitsevent in up to 40 recent files under$CODEX_HOME/sessions/**/*.jsonl(~/.codexby default). File modification time never stands in for measurement time. A malformed trailing line does not erase the last valid event. A newer empty measurement remains unknown.
What ships with it
1 file 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.
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.
- 26d ago First seen · 86 lines · 50 tokens per session scan A 2f15ce729ce5
limit is a skill published in the GitHub repository avenoxai/avenoxskills (93 stars, last pushed 22d ago), licensed MIT. It adds 50 tokens to every session and 1,135 once invoked, about $0.0002 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 2026-09-22.
Other skills, from other repositories
cron
Use this skill only for scheduled or recurring tasks. Manage jobs with qwenpaw cron list/create/get/state/update/pause/resume/delete/run, and always pass --agent-id explicitly.
mailbox
A guide for performing email and mailbox tasks through the QwenPaw mailbox system. It covers supported personal email providers, credentials, configuration, and the required mailbox workflow.
cron
A procedure for creating and managing tasks that run automatically at a future time or repeat on a schedule. It uses QwenPaw's scheduling commands and requires an agent identifier.
channel_message
Use this skill to proactively send a one-way message to a user/session/channel, usually only when the user explicitly asks to send to a channel/session or when proactive notification is needed. First query sessions with qwenpaw chats list, then push with qwenpaw channels send.
wiki-dashboard
Create dynamic, queryable Obsidian dashboard views using Bases or Dataview. Use for auto-updating tables, indexes, or structured views of vault content; not for ordinary question answering.
wiki-switch
List, create, or switch named Obsidian wiki vault profiles under the global config directory. Use when changing the active vault, checking which vault is active, or managing multi-vault profiles.