agentleFS
Sign inStart free
MGrinMGrin/bb-plugin-accounts/skills/accounts/SKILL.md

accounts

Skill4 starsChanged 2 months agoMITView on GitHub

Read custom Claude and Codex subscription telemetry, switch captured Claude/Codex accounts, or see which accounts mgrin has paused in the accounts plugin.

5 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use content and research.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Rayan55050Skill

    content

    Создание контента: посты, статьи, тексты

    Also in 3 projects that use this

  • NVIDIASkill

    research

    Before designing a feature, research how comparable libraries solve the problem, check best practices, known pitfalls, and upstream changes. Use when starting a new feature, investigating an unfamiliar area, or the user asks to research something.

    Also in 3 projects that use this

  • Other versions5 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 5

    1. notque/claude-code-toolkit★ 448same name
    2. notque/vexjoy-agent★ 448same name
    3. qbeka/jev-job-search★ 5same name
    4. harshsinghmp/muse-skills★ 1same name
    5. grechj/Australian-Builders-Brain★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository2 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use email or label for display, never accountId.

    line 16
  2. Very specific:

    Read paused before reading capacity: it is a decision, not a fault.

    line 54
  3. Distinctive:

    Not to save quota, not to dodge a limit, not because an account looks unhealthy.

    line 58

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Custom subscription accounts

bb accounts telemetry [--json] returns normalized provider observations. Version 1 has accounts[] and separate thread tokens[]. Read providerId, scope, observedAt, fresh and capacity together. A successful read exits 0 even when capacity is UNKNOWN; this is a reporting command, not an admission gate.

  • capacity is subscription headroom only: available, exhausted, unavailable or unknown. Credits and token counts never make a subscription available.
  • Codex's main codex bucket controls ordinary capacity. Spark is shown separately.
  • Use email or label for display, never accountId. codex_account_email comes from read-only account/read (refreshToken: false); missing email stays unavailable and does not invalidate quota. The UI defaults to usage/reset, with diagnostics collapsed.
  • A Codex local-session observation without accountId cannot identify an account. SDK quota observations stay thread-scoped and cannot replace the current local snapshot.
  • Missing, malformed, stale, future-dated or incomplete main quota is UNKNOWN. A reset passing is not evidence of recovery: read a fresh observation.
  • Codex account rows read the private ~/.config/codex-usage/usage.json slot cache (256 KiB max). The dotfiles usage-only poller gathers native app-server quota in isolated credential homes every 180 seconds. Stale rows are UNKNOWN. slot and active identify the live login on disk, not every running process.

The legacy bb accounts list, outage, place, auto, switch, log, stats and forecast commands remain Claude-only. In particular, outage includes Claude paid credits by design; it is not the new subscription-only capacity field.

bb accounts switch <slot> changes live Claude credentials. bb accounts codex use <slot> changes the live Codex login through the dotfiles codex-acct helper with expected-current identity protection. Capture logins using codex-acct capture <slot> after an official human sign-in. Never read or print auth blobs. bb accounts codex auto runs one decision; codexAutoSwitch defaults off. Enable only after running-session refresh behavior has been verified. codexSwitchAt defaults to 97; candidates require fresh free main windows and the helper rechecks capacity during switching. Cooldown is 120 seconds. Spark and credits cannot make a candidate eligible.

The usage page includes native Codex response tokens by model/repo/directory. Input includes cache; output includes reasoning; neither maps to quota percent. Missing turn context is shown as unknown/unresolved. Codex forecasts use each account/window's own measured quota history; they remain provisional until three days, 1,440 polls, and 1,296 valid adjacent intervals. Forecasts stop at reset and do not predict post-reset use.

Paused accounts

bb accounts pause <slot> / bb accounts resume <slot> (Claude) and bb accounts codex pause <slot> / bb accounts codex resume <slot> (Codex) pause or resume an account. Paused means no request of any kind is made on it — it is not polled, nothing switches or places work onto it, and switch / codex use refuse it naming resume. It stays listed everywhere, marked PAUSED, with its last-known usage and that reading's age. It is not disabled: resume is one command. The active account cannot be paused; switch away first.

  • In telemetry --json a paused row has paused: {at}, capacity unknown and fresh false. Read paused before reading capacity: it is a decision, not a fault. Never report a paused account as unreadable, blind, lost or as missing capacity, and never raise an alarm about one.
  • A paused account's windows are LAST-KNOWN: say how old they are, never present them as current.
  • Pausing and resuming are mgrin's acts: agents never pause or resume an account on their own. Not to save quota, not to dodge a limit, not because an account looks unhealthy. If a pause seems wrong, tell mgrin; do not undo it.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free