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 burin-labs/harn --skill harn-control-eventsgit clone --depth 1 https://github.com/burin-labs/harnWrote 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/burin-labs/harn/harn-control-events)<a href="https://agentmods.dev/skills/burin-labs/harn/harn-control-events"><img src="https://agentmods.dev/badge/skills/burin-labs/harn/harn-control-events/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/burin-labs/harn/harn-control-events"><img src="https://agentmods.dev/badge/skills/burin-labs/harn/harn-control-events.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00029 | $0.01285 |
| Opus 5 | $0.00015 | $0.00642 |
| Sonnet 5 | $0.00006 | $0.00257 |
| Haiku 4.5 | $0.00003 | $0.00128 |
Grade A, and why
harn-control-events 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 11d 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 — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Harn control events
A control event is a person telling a running session to stop or to change course. It is an event the session must answer, not a request it may finish its current plan first and then consider.
Pair this skill with [[harn-orchestration]] for where the loop drains injections, [[harn-agent]] for loop and terminal vocabulary, and [[harn-product-quality]] for what a host must show while a control is in flight.
The three things a stop has to do
A stop that does one or two of these is not a stop. Check all three separately, because the second and third fail silently while the first looks correct.
- The loop stops. No further iteration, model call, or tool dispatch after the control lands.
- What the session started dies. A cancelled agent that leaves a backgrounded command running has moved the work, not stopped it. Kill the handles owned by that session and no others.
- The record names the control. The run's terminal says a person stopped it. Without this, a stopped run and a run that merely produced nothing are the same row in every downstream report.
Point 3 is the one that rots. An accepted cancel is answered — the prompt
returns stopReason: "cancelled" — so the turn settles normally and any
tagging that fires only when a turn fails to settle never fires at all. Seal
the terminal from the observed control, not from whether the turn looked
finished.
Absence must not read as success
The dangerous shape is a control that names something that is not there.
- A cancel naming an unregistered session must be loud: warn, and emit a
ControlOutcomewithstatus: "rejected"naming the session. Consuming the frame silently makes "nothing was cancelled" indistinguishable from "everything was cancelled". - Every control decision, accepted or rejected, gets a
ControlOutcomecarrying method, outcome, status, actor, target, and reason. Read the structured record; do not grep prose for the verdict. - A record that a control arrived is not a record that the loop stopped. Assert the loop counters after the control, not the frame before it.
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.
- 11d ago First seen · 110 lines · 29 tokens per session scan A a8c4037fc77a
harn-control-events is a skill published in the GitHub repository burin-labs/harn (21 stars, last pushed yesterday), licensed Apache-2.0. It adds 29 tokens to every session and 1,285 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 skills, from other repositories
translate
A translation tool for converting user-provided text between languages, with Chinese and English as the default pair.
agent-communication-protocol
Open protocol for AI agent interoperability enabling standardized communication between agents, applications, and humans across different frameworks.
self-host-studio
Install and self-host deco Studio (the open-source control plane) on the user's own infrastructure. Use when the user wants to install, run, self-host, or deploy Studio locally (Docker, Rancher Desktop, kind, minikube) or on Kubernetes (managed or self-managed), or asks to "install Studio", "self-host decocms"…
gh-issue
Fetch a GitHub issue, create a branch, implement with TDD, and open a PR.
gh-issues
Walk over all open GitHub issues that are unassigned or assigned to the current user, and process each one via the /gh-issue skill, sequentially.
module-new
Scaffold a new Gacela module under src/php/ with Facade, Provider, and CLAUDE.md.