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 commands/cputester5465/countbean-plugin/connectgit clone --depth 1 https://github.com/CPUtester5465/countbean-pluginWrote 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/commands/cputester5465/countbean-plugin/connect)<a href="https://agentmods.dev/commands/cputester5465/countbean-plugin/connect"><img src="https://agentmods.dev/badge/commands/cputester5465/countbean-plugin/connect.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 | $0.00023 | $0.00691 |
| Opus 5 | $0.00012 | $0.00345 |
| Sonnet 5 | $0.00005 | $0.00138 |
| Haiku 4.5 | $0.00002 | $0.00069 |
Grade A, and why
connect 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 3d 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.
What it actually says
Connect this Claude session to the user's hosted Countbean book.
Arguments (may arrive in either order, on one line, or not at all): $ARGUMENTS
Do this in order:
-
If
$ARGUMENTScontains acbk_…token, callconnect_bookwith it (pass the whole argument string asapi_key— the tool sortscbk_…frombok_…itself). Then go to step 4. -
If there are no arguments, call
connection_statusfirst.-
Already connected and reachable → say which book, and stop. Do not reconnect.
-
Not connected → prefer browser approval. It needs no key at all, so do not ask for one. Two calls, in this order: a.
start_device_authorization— returns immediately with a short code and a link. b. Show the user both, verbatim, and wait for them to approve. They open the link, pick which book to connect, and approve it. c.await_device_approval— blocks until they approve, then saves the connection.Never call (c) before showing the code from (a): until the user sees it they have nothing to type, and the grant expires in ten minutes.
-
Only if the user would rather paste a key by hand — or their browser is on a different machine — tell them where it is, in this exact shape:
Open your book at app.countbean.com → Connect Claude → Create key. It shows you one line to paste back here. The key is shown once. Then stop and wait. Do not guess a key, do not offer to create one — the plugin cannot mint keys, only the signed-in book page can.
-
-
If the user pastes something that is not a key (e.g. only a
bok_…), say which half is missing and where it is shown. Never write a partial connection. -
Report the result verbatim-ish.
connect_bookverifies against the live book before it saves, so its answer is authoritative:- Connected → confirm the book id, then call
book_statusand show the account count and latest commit, so the user sees their real book rather than a claim about it. - Refused → relay the reason. The common ones are a mistyped paste, a key that was revoked, and a key issued for a different book than the id given. Do not retry silently.
- Connected → confirm the book id, then call
Notes:
- This takes effect immediately — no restart, no environment variables, no
.envedit. Credentials are stored in~/.countbean/credentials.json, readable only by the user. - If the user would rather use a
.envor shell exports, those still work and take precedence:COUNTBEAN_API_KEYandCOUNTBEAN_BOOK_ID. Mention this only if they ask. - To undo:
disconnect_book.
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.
- 3d ago First seen · 49 lines · 23 tokens per session scan A ab6f0f89afc4
connect is a command published in the GitHub repository CPUtester5465/countbean-plugin (17 stars, last pushed 8d ago), licensed MIT. It adds 23 tokens to every session and 691 once invoked, about $0.0001 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 2026-08-30.
Other commands, from other repositories
report
Command "report" from rustledger/rustledger, covering rledger report, usage, subcommands, global options and examples.
price
Command "price" from rustledger/rustledger, covering title: rledger price description: fetch commodity prices, rledger price, usage, arguments and options.
add
Command "add" from rustledger/rustledger, covering rledger add, usage, arguments, options and interactive mode.
format
Command "format" from rustledger/rustledger, covering rledger format, usage, arguments, options and examples.
index
Command "index" from rustledger/rustledger, covering commands reference, commands, global options, specifying the ledger file and explicit file path.
review
Kernel first (/ledger-verify or /ledger-audit), then host TDD/review/security if present. See skills/ledger-review/SKILL.md.