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.
git clone --depth 1 https://github.com/datacore-one/datacoreWrote 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/datacore-one/datacore/process-inbox)<a href="https://agentmods.dev/commands/datacore-one/datacore/process-inbox"><img src="https://agentmods.dev/badge/commands/datacore-one/datacore/process-inbox/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/commands/datacore-one/datacore/process-inbox"><img src="https://agentmods.dev/badge/commands/datacore-one/datacore/process-inbox.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 | $0.00008 | $0.01783 |
| Opus 5.5 | $0.00003 | $0.00713 |
| Sonnet 5.5 | $0.00002 | $0.00357 |
| Haiku 4.5 | $0.00001 | $0.00178 |
Grade A, and why
process-inbox 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 19d 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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/process-inbox
Automated inbox processing — classify and route all entries from inbox.org.
When to Use
- The nightly inbox job (chief-of-staff
cos_inbox.sh), once per space with an inbox - Manual batch processing when inbox is full
- After email triage has added new items
Mode Detection
Nightshift mode (no user present — the caller says so, e.g. "nobody is present", "unattended", "nightshift mode"): Process all items autonomously. Make classification decisions without asking. Never delete an entry. An entry you cannot decide is NOT guessed into a list: it stays in the inbox, marked [NEEDS_REVIEW] (see "Entries you cannot decide" below), for the owner to review.
Interactive mode (user present): Ask for confirmation on ambiguous items. If the owner does not decide one, it stays in the inbox marked [NEEDS_REVIEW], the same as at night.
The three outcomes — every entry ends in exactly one
- Clarified and moved to where it belongs (the routing rules below). It leaves the inbox.
- Finished — its state is DONE or CANCELLED. It leaves the inbox and is kept: move the whole entry, unchanged (state,
CLOSED:stamp, properties, body), to the end of[space]/org/inbox_archive.org(create it with#+TITLE: Inbox archiveif missing). Never reopen it (no TODO/NEXT), never route it into next_actions.org as an open task, never delete it. - Cannot decide — you cannot tell what it is or what the next action would be (a fragment such as "blue thing w/ Marko??"). It STAYS in the inbox, where it is, with the literal text
[NEEDS_REVIEW]put at the front of its title, after any state keyword:** TODO [NEEDS_REVIEW] blue thing w/ Marko??— a prefix in the title, not a:NEEDS_REVIEW:tag and not a state. Not someday.org, not next_actions.org, not "Operations with a note". An entry already marked[NEEDS_REVIEW]is left as it is.
Nothing else remains in the inbox after a run.
Workflow
Step 1: Read inbox.org
Read the inbox you were given ([space]/org/inbox.org). When none is named, use the personal space's inbox: the space install.yaml names as roles.personal (never assume a folder name or number).
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.
- 19d ago Changed · +18 lines 4043754a50cd
- 21d ago Changed · +1 lines 294684d8e2aa
- 1mo ago First seen · 104 lines · 8 tokens per session scan A ff5b6b6b2496
process-inbox is a command published in the GitHub repository datacore-one/datacore (4 stars, last pushed yesterday), licensed MIT. It adds 8 tokens to every session and 1,783 once invoked, about $0.0000 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 2026-09-03.
Other commands, from other repositories
linear-drain
Create the queued Linear projects and issues that shell scripts captured offline.
loops
Weekly review — surface everything in-flight and force a ship / block / kill decision on each.
open
Open the memex vault in Finder or Obsidian.
telegram-secretary
A command entry point for registering, configuring, and managing a Telegram secretary that runs continuously on Anthropic’s cloud routines. Its specification is kept in a referenced skill file and routine prompt.
para-status
Show an overview of your PARA system — item counts, statuses, and attention-needed flags.
dejavu-resume
Find the best matching past Claude Code session and produce a ready-to-run resume command.