hor-agent-loop SKILL.md skillA

hor-agent-loop from openreachtech/hora-skills-ort-renchan is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add openreachtech/hora-skills-ort-renchan/hor-agent-loop@git:20260905.5a04969

Pins git:20260905.5a04969 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
openreachtech/hora-skills-ort-renchan
Path
kit/skills/hor-agent-loop/SKILL.md
Versions recorded
1
Latest version
git:20260905.5a04969, committed Sep 5, 2026
Size
8,018 bytes
Repository stars
32
License
Apache-2.0

Build an LLM agent loop (iterative agent) in a backend repository with the three @openreachtech/mentsu-agent-loop packages — core (the transport-agnostic iteration engine), renchan-job (run a loop as a BullMQ/Redis job with progress publishing), and graphql (a mutation that launches the loop plus a subscription that streams its progress). Use whenever the user asks to add or edit an agent loop, an agent action, a composite agent, or to stream agent progress.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Agent Loop (building agent loops with mentsu-agent-loop)

A skill for implementing an LLM agent's iteration ("loop") in a backend. The iteration engine,
progress delivery, composition, job execution, and GraphQL integration are all provided by the
**three `@openreachtech/mentsu-agent-loop-*` packages**. The app writes only the domain-specific
**actions and loops**; iteration, stopping, compression, parallelism, and transport ride on the
ready-made base classes.

This skill covers the libraries' specs and use cases, plus the integration patterns for renchan's
GraphQL and renchan-job-bullmq. **Always use these three packages** (never hand-roll your own
iteration engine, queue bridge, or progress channel). The details are split into the [detail
files](#detail-files) at the end.

## Grand principle: write actions and loops without knowing about AI, DB, queues, or GraphQL (one-way dependency)

The center of this package family's design is a **one-way dependency direction (outer → core)**.
The iteration logic the app writes (actions / loops) **imports no business dependency and no
transport**; everything is reached through the **`context` (an opaque DI seam)** and through **thin
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add openreachtech/hora-skills-ort-renchan/hor-agent-loop@sha256:dbace9e4bc72b7fc

The line at the top of this page pins the label git:20260905.5a04969, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_qvh42yrcetiwofq5.svg)](https://markdownregistry.com/a/art_qvh42yrcetiwofq5)

2 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 1 version of this file, committed on Sep 5, 2026.

versioncommittedcommitsizeaudit
git:20260905.5a04969 latest2026-09-05 5a04969 8,018 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (8018 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

openreachtech/hora-skills-ort-renchan · 32 stars · license Apache-2.0 · pushed 2026-10-16 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_qvh42yrcetiwofq5
GET https://markdownregistry.com/api/v1/resolve?ref=openreachtech/hora-skills-ort-renchan/hor-agent-loop
GET https://markdownregistry.com/api/v1/blob/dbace9e4bc72b7fcf13fe1d312215a26c379fb71f75c0777a571e4081ad07fab

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from openreachtech/hora-skills-ort-renchan

audit SKILL.md skill
openreachtech/hora-skills-ort-renchan · .claude/skills/audit/SKILL.md · Repository-specific check that every skill sits directly under kit/skills/, carries its `hor-` prefix, and declares a…
git:20260906.411ca02 · audit A · 32 stars
CLAUDE.md CLAUDE.md file
openreachtech/hora-skills-ort-renchan · CLAUDE.md
git:20260918.ad78484 · audit A · 32 stars
hor-ai-agent-structure SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-ai-agent-structure/SKILL.md · Structure an app-side AI agent on @openreachtech/mentsu-agent-loop-core in a renchan backend: an app/agents/<name>/…
git:20260905.5a04969 · audit A · 32 stars
hor-ai-prompt-document-store SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-ai-prompt-document-store/SKILL.md · Store and compose AI-agent prompts and documents from the database rather than hard-coding them in a renchan backend —…
git:20260905.5a04969 · audit A · 32 stars
hor-backend-testing SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-backend-testing/SKILL.md · Organize and run the backend repository's tests: where a test file goes (tests/__tests__ for tests that do not write to…
git:20261016.ade0c6a · audit A · 32 stars
hor-bank-id SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-bank-id/SKILL.md · Allocate an exclusive, collision-free row-id prefix for a requester inside one backend repository, so two writers never…
git:20260905.5a04969 · audit A · 32 stars
hor-build-e2e-test-environment SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-build-e2e-test-environment/SKILL.md · Build, run and debug the manual local E2E environment under `e2e/docker/`, a compose stack of the product's real…
git:20261016.a9876a7 · audit B · 32 stars
hor-constant-definition SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-constant-definition/SKILL.md · Define application constants the Renchan way: one constant (one category) per file, ALWAYS as two files — a CommonJS…
git:20260905.5a04969 · audit A · 32 stars
hor-cookie-authentication SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-cookie-authentication/SKILL.md · Password sign-in with cookie-based sessions for a renchan backend, per actor: the credential and token models (password…
v2026.08.18 · audit A · 32 stars
hor-database-design SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-database-design/SKILL.md · Decide the logical shape of a renchan database schema — the design choices made before writing a migration or model…
git:20260905.5a04969 · audit A · 32 stars
hor-execution-placement-pattern SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-execution-placement-pattern/SKILL.md · Decide where a piece of processing (especially a write / state change) should be implemented in a web system: as a…
git:20260905.5a04969 · audit A · 32 stars
hor-external-api-client SKILL.md skill
openreachtech/hora-skills-ort-renchan · kit/skills/hor-external-api-client/SKILL.md · Implement a client for an external HTTP/REST API using `@openreachtech/mentsu-rocket-client` (the Launcher / Payload /…
git:20260905.5a04969 · audit A · 32 stars

Every file in openreachtech/hora-skills-ort-renchan

Browse by kind, by grade A, or by owner.