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 azeemkafridi/bulkpublish-api --skill inbox-triagegit clone --depth 1 https://github.com/azeemkafridi/bulkpublish-apiWrote 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/azeemkafridi/bulkpublish-api/inbox-triage)<a href="https://agentmods.dev/skills/azeemkafridi/bulkpublish-api/inbox-triage"><img src="https://agentmods.dev/badge/skills/azeemkafridi/bulkpublish-api/inbox-triage/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/azeemkafridi/bulkpublish-api/inbox-triage"><img src="https://agentmods.dev/badge/skills/azeemkafridi/bulkpublish-api/inbox-triage.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 | $0.00064 | $0.00923 |
| Opus 5.5 | $0.00026 | $0.00369 |
| Sonnet 5.5 | $0.00013 | $0.00185 |
| Haiku 4.5 | $0.00006 | $0.00092 |
Grade A, and why
inbox-triage 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 — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Inbox Triage
The Inbox gathers direct messages, reviews and comment threads from the user's connected accounts, on the networks that allow it. This skill reads it, sorts it and helps the user answer.
The rule that matters most
Every reply goes to a real person, immediately, and cannot be unsent. Before calling reply_to_conversation:
- Show the user the exact text you will send.
- Name the recipient: the network, the user's account it goes out from, and the person it is addressed to.
- Wait for an explicit yes for that message. Approval of one reply is not approval of the next.
Never send in bulk. If the user asks you to answer everything, draft every reply, show them as a numbered list, and send only the ones the user confirms. Drafting is always safe; sending is not.
Tools
list_conversations: open, archived or snoozed conversations. Filters:status,kind(dm,review,comment),platforms,channelId,assigned(me,unassignedor a teammate's id),qfor text search,limit, andcursorfrom the previous page.get_conversation: the messages in one conversation, oldest first. Reading it does not mark it read.update_conversation: mark read or unread, archive or reopen, snooze until a time (at most 90 days ahead), assign to a teammate or unassign. Every change can be undone with another call.reply_to_conversation: sendstext(up to 4,000 characters). In a direct message it can attach up to 4mediaIdsfromupload_media. In a comment thread it answers the newest comment from the other person, or the one given asreplyToMessageId.
Recipes
What is waiting for me
list_conversationswithstatus: "open", addingassigned: "me"if the user asks about their own.get_conversationon the ones that look urgent: complaints, questions with a deadline, low-star reviews.- Summarise each in one line: who wrote, on which account, what they want, how long ago.
- Offer next steps (reply, assign, snooze, archive) and act only on what the user picks.
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 · 56 lines · 64 tokens per session scan A 634390c8e326
inbox-triage is a skill published in the GitHub repository azeemkafridi/bulkpublish-api (3 stars, last pushed yesterday), licensed MIT. It adds 64 tokens to every session and 923 once invoked, about $0.0003 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 2026-10-21.
Other skills, from other repositories
skill-profile-manager
A profile-management workflow for creating, editing, switching, exporting, comparing, and updating creator profiles. A profile is a structured record of an account’s identity, style, audience, platforms, preferences, and accumulated notes.
template-library
A personal library for saving, reusing, listing, editing, and versioning content templates. Templates can include replaceable fields such as a title, problem, solution, or call to action.
skill-my-account
A read-only account lookup workflow for checking the user’s own logged-in Xiaohongshu, Douyin, Kuaishou, Zhihu, or WeChat Channels accounts. It can show account identity and available creator statistics such as followers, likes, following, works, and post lists.
skill-publish-checklist
A quick checklist for confirming that a piece of content contains the required publishing parts, such as a title, cover, tags, format, link, and call to action.
skill-publish-scheduler
A scheduling tool for organising content by platform and publish time, then finding items that are due and handing them to the appropriate publishing tool.
skill-publish-notify
A notification tool that sends publishing results to team chat services or any webhook. A webhook is a URL that accepts an automated message from another tool.