mention-gate

mention-gate is a skill for Claude Code, Codex from sonichi/sutando. It costs 65 tokens per session (1,679 once invoked), scanned A, original, MIT.

A message filter setting that decides whether messages tagging the owner, but not the bot, count as bot mentions in restricted channels.

In plain words
What is it for?
Use it to admit owner-tagged messages as tasks in channels that require a bot mention, without changing later access checks or unrestricted channels.
Why use it?
It lets those messages reach the bot when enabled, while keeping the current blocking behavior by default. Accepted messages are recorded in an audit log for review.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/sonichi/sutando/mention-gate
Any agent
npx skills add sonichi/sutando --skill mention-gate
Clone the repo
git clone --depth 1 https://github.com/sonichi/sutando

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for mention-gate

README.md
[![agentmods](https://agentmods.dev/badge/skills/sonichi/sutando/mention-gate.svg)](https://agentmods.dev/skills/sonichi/sutando/mention-gate)
Your own site
<a href="https://agentmods.dev/skills/sonichi/sutando/mention-gate"><img src="https://agentmods.dev/badge/skills/sonichi/sutando/mention-gate.svg" alt="Measured on agentmods" height="20"></a>
Per session 65 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,679 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00065 $0.01679
Opus 5 $0.00032 $0.00839
Sonnet 5 $0.00013 $0.00336
Haiku 4.5 $0.00006 $0.00168

Measured 6d ago against content hash ac0cd1da638a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

mention-gate 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.

The scan reads SKILL.md. This mod also ships 2 executable files (scripts/mention-gate.py, scripts/witness.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

skills/mention-gate/SKILL.md · 122 lines

How it starts

The opening of the file, as written. The whole thing — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Mention Gate

Today, in every requireMention channel, a message @-tagging the owner (but not the bot) is never ingested — it stays invisible to the fleet. This skill adds the missing ON side:

  • ON — an owner-tagged message from anyone but the owner counts as a bot mention: it passes the requireMention check and becomes a task. Every message pulled in this way is appended to a durable audit log so the owner can review exactly what the toggle admitted.
  • OFF (default) — today's behavior, unchanged.

Never affected: messages from the owner himself (they follow the existing rules), messages that actually mention the bot (ingested in both states, as always), and requireMention:false channels (everything there already ingests, gate on or off). Later access gates (channel allowFrom, tier resolution) still apply to an admitted message — the gate substitutes only the missing bot-mention, nothing else.

Fail-closed: the feature ADDS ingestion, so every failure reads as OFF — missing or malformed state, an unparseable until, or any exception inside the bridge's gate check leaves the ordinary requireMention rejection standing. A broken state file must never surprise-ingest.

Usage

python3 skills/mention-gate/scripts/mention-gate.py on              # owner tags trigger ingestion
python3 skills/mention-gate/scripts/mention-gate.py on --for 2h     # with auto-expiry back to off (30m/2h/1d)
python3 skills/mention-gate/scripts/mention-gate.py off             # today's behavior (default)
python3 skills/mention-gate/scripts/mention-gate.py status          # state + how many messages pulled in

State + audit

Scope: PER-HOST, not fleet-wide

The gate is stored under <workspace>/state/, which vault.sync.exclude lists explicitly — the vault runs in whitelist mode and carries only notes/, talks/, hosts/ and the memory tree. git ls-files state/mention-gate.json returns nothing on a live host. So toggling the gate changes only the host it was run on; every other host keeps its own setting. Run the command once per host. There is currently no synced fleet-policy surface to store it on.

Read the full file on GitHub · 122 lines

Files

What ships with it

2 files 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.

Changes

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.

  1. 6d ago First seen · 122 lines · 65 tokens per session scan A ac0cd1da638a

Subscribe to this mod's changes

mention-gate is a skill published in the GitHub repository sonichi/sutando (391 stars, last pushed today), licensed MIT. It adds 65 tokens to every session and 1,679 once invoked, about $0.0003 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.