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 resonatehq/resonate-skills --skill resonate-human-in-the-loop-pattern-rustgit clone --depth 1 https://github.com/resonatehq/resonate-skillsWrote 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/resonatehq/resonate-skills/resonate-human-in-the-loop-pattern-rust)<a href="https://agentmods.dev/skills/resonatehq/resonate-skills/resonate-human-in-the-loop-pattern-rust"><img src="https://agentmods.dev/badge/skills/resonatehq/resonate-skills/resonate-human-in-the-loop-pattern-rust/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/resonatehq/resonate-skills/resonate-human-in-the-loop-pattern-rust"><img src="https://agentmods.dev/badge/skills/resonatehq/resonate-skills/resonate-human-in-the-loop-pattern-rust.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 | $0.00000 | $0.02377 |
| Opus 5 | $0.00000 | $0.01189 |
| Sonnet 5 | $0.00000 | $0.00475 |
| Haiku 4.5 | $0.00000 | $0.00238 |
Grade A, and why
resonate-human-in-the-loop-pattern-rust 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 12d 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 — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Resonate Human-in-the-Loop Pattern — Rust
SDK note (0.6.0).
ctx.promise::<T>()is a realpub fnin the Rust SDK source (resonate-sdk-rs:resonate/src/context.rs) with a fullPromiseTask<T>builder, but it is not yet covered indocs/develop/rust.mdx. The API is safe to use; cite source paths when reviewers ask. APIs may shift between 0.x releases.
Overview
A human-in-the-loop workflow blocks a durable function on a promise that an external actor settles — a reviewer clicking "approve," a webhook firing from a third-party system, an operator running a CLI command. The worker doesn't poll; it awaits the PromiseTask future and Resonate parks the execution until the promise settles.
For the language-agnostic mental model, see resonate-human-in-the-loop-pattern-typescript. The Rust shape differs in (a) type-parameterized promises ctx.promise::<T>(), (b) lazy builder with .timeout / .data / .id / .create / .await, (c) SDK-generated IDs you fetch via .id().await?.
When to use
- Approval gates (expense, deploy, content moderation)
- Third-party callbacks (Stripe, DocuSign, Twilio)
- Operator unblock steps (break-glass in runbooks)
- Any step where the data or decision comes from outside the Resonate worker set
Basic shape
use resonate::prelude::*;
use serde::{Serialize, Deserialize};
use std::time::Duration;
#[derive(Serialize, Deserialize)]
struct Decision {
approved: bool,
reviewer: String,
}
#[resonate::function]
async fn expense_approval(ctx: &Context, expense_id: String) -> Result<String> {
// build a promise with a 24-hour SLA and the expense ID attached as data
let task = ctx
.promise::<Decision>()
.timeout(Duration::from_secs(24 * 60 * 60))
.data(&serde_json::json!({ "expense_id": expense_id }))?;
// fetch the SDK-generated promise ID so an external actor can target it
let promise_id = task.id().await?;
// stash the ID where the reviewer UI / webhook will see it
ctx.run(save_approval_id, (expense_id.clone(), promise_id.clone())).await?;
// block until someone resolves/rejects/cancels the promise
let decision: Decision = task.await?;
if decision.approved {
ctx.run(process_reimbursement, expense_id.clone()).await?;
Ok(format!("approved by {}", decision.reviewer))
} else {
Ok(format!("rejected"))
}
}
async fn save_approval_id((expense_id, promise_id): (String, String)) -> Result<()> {
// write (expense_id, promise_id) to a DB or queue a notification
Ok(())
}
async fn process_reimbursement(expense_id: String) -> Result<()> {
Ok(())
}
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.
- 12d ago First seen · 225 lines · 0 tokens per session scan A 55ac20cfe468
resonate-human-in-the-loop-pattern-rust is a skill published in the GitHub repository resonatehq/resonate-skills (6 stars, last pushed 20d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 2,377 tokens. 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 2026-08-31.
Other skills, from other repositories
golem-make-http-request-rust
Making outgoing HTTP requests from a Rust Golem agent. Use when the user asks to call an external API, make HTTP requests, use an HTTP client, or send HTTP requests from agent code.
golem-add-llm-rust
Adding LLM and AI capabilities to a Rust Golem agent. Use when the user wants to add LLM chat, embeddings, web search, vector DB, graph DB, document search, video generation, speech-to-text, text-to-speech, or any AI provider integration.
golem-parallel-workers-rust
Fan out work to multiple parallel agents and collect results in a Rust Golem project. Use when the user asks about parallel execution, fan-out/fan-in, spawning child agents for parallel work, forking, or aggregating results from multiple agents.
golem-retry-policies-rust
Configuring semantic retry policies for a Rust Golem agent. Use when the user asks about retry policies, retry strategies, exponential backoff, error handling retries, transient error recovery, retry predicates, withRetryPolicy, withnamedpolicy, NamedPolicy, Policy composition, jitter, countBox, timeBox, andThen, or…
golem-add-http-endpoint-rust
Exposing a Rust Golem agent over HTTP. Use when the user asks to add HTTP endpoints, mount an agent to a URL path, or expose agent methods as a REST API.
golem-atomic-block-rust
Using atomic blocks, idempotency, and oplog management in a Rust Golem project. Use when the user asks about atomically, idempotence mode, oplog commit, or idempotency keys.