---
type: "post"
id: "an-author-on-habr-describes-building-aiworks-a-b51ea4"
slug: "an-author-on-habr-describes-building-aiworks-a-b51ea4"
title: "An author on Habr describes building AIWorks, a local supervisor for Claude Code"
description: "The thesis is that unattended agent work depends less on asking a model to continue than on defining a bounded loop with verifiable outputs and a human-controlled merge boundary."
retrieval_nugget: "The thesis is that unattended agent work depends less on asking a model to continue than on defining a bounded loop with verifiable outputs and a human-controlled merge boundary."
published_at: "2026-10-15"
updated_at: "2026-10-15"
record_date: "2026-10-15"
date_kind: "published_at"
topics: ["claude","claude-code"]
source_urls: ["https://habr.com/ru/articles/1094148?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:an-author-on-habr-describes-building-aiworks-a-b51ea4"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:7ee9b080ac336d48edca6362c81d0df265a76957f7b49d2baccd3e38a9cdfd3b","reviewed_at":"2026-10-15T10:05:19.345Z","source_evidence_count":1,"verified_claim_count":3,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":3,"implication_count":2,"watch_condition_count":1,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"The thesis is that unattended agent work depends less on asking a model to continue than on defining a bounded loop with verifiable outputs and a human-controlled merge boundary.","observed_facts":[{"text":"The author reports AIWorks ran on its own repository for 24.5 hours and produced 63 completed pull requests, each independently checked.","source_urls":["https://habr.com/ru/articles/1094148?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]},{"text":"The author reports up to four concurrent workstreams and 18 pull requests merged between midnight and 7 a.m.","source_urls":["https://habr.com/ru/articles/1094148?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]},{"text":"The described system used a fresh read-only session to check the same commit and rerun tests; it opened pull requests but did not merge or deploy.","source_urls":["https://habr.com/ru/articles/1094148?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]}],"mechanism":"The mechanism is separation of duties and constrained authority. The author says a fresh, read-only session checks the same commit and reruns tests; the supervisor pushes that verified commit and opens a pull request, but does not merge or deploy.","why_now":"Why now: coding agents increasingly invite longer unattended runs, making the cost of weak verification and ambiguous completion more visible.","implications":["The first implication is that teams extending agent autonomy should define what artifact counts as complete and keep irreversible actions behind a separate approval boundary.","Pull-request volume is an incomplete proxy for useful output without review quality and rework measures."],"evidence_boundary":"The evidence boundary is one self-reported case, and the article does not establish that the workflow generalizes or that every reported change had equal value.","watch_conditions":["The first watch condition is whether subsequent projects report independently reproducible outcomes, including review quality and rework, rather than pull-request volume alone."],"related_records":[],"new_branch_reason":"No prior New Runtime analytical record was supplied; this case merits a distinct record about verification and authority boundaries in unattended coding-agent loops."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/an-author-on-habr-describes-building-aiworks-a-b51ea4/","markdown":"https://newruntime.com/posts/an-author-on-habr-describes-building-aiworks-a-b51ea4.md","json":"https://newruntime.com/posts/an-author-on-habr-describes-building-aiworks-a-b51ea4.json"}
related_patterns: ["goal-scoped-loops-replace-manual-continuation"]
related_shifts: ["goal-scoped-agent-loops"]
---

# An author on Habr describes building AIWorks, a local supervisor for Claude Code

## Retrieval answer

The thesis is that unattended agent work depends less on asking a model to continue than on defining a bounded loop with verifiable outputs and a human-controlled merge boundary.

The thesis is that unattended agent work depends less on asking a model to continue than on defining a bounded loop with verifiable outputs and a human-controlled merge boundary. The Habr author reports that AIWorks ran on its own repository for 24.5 hours and produced 63 completed pull requests, each independently checked. At peak, four workstreams ran in parallel; 18 pull requests were merged overnight. These figures describe one author's project, not a controlled productivity comparison.

The mechanism is separation of duties and constrained authority. The author says a fresh, read-only session checks the same commit and reruns tests; the supervisor pushes that verified commit and opens a pull request, but does not merge or deploy. That structure makes a checkable artifact—not an agent's assertion of completion—the handoff. The author also reports exhausting GitHub GraphQL quota and discovering a test that passed on Windows but failed on Linux because of filesystem case sensitivity. These failures show why repeatability and explicit publication semantics are part of the loop, not incidental cleanup.

Why now: coding agents increasingly invite longer unattended runs, making the cost of weak verification and ambiguous completion more visible. The evidence boundary is one self-reported case, and the article does not establish that the workflow generalizes or that every reported change had equal value. The first implication is that teams extending agent autonomy should define what artifact counts as complete and keep irreversible actions behind a separate approval boundary. The first watch condition is whether subsequent projects report independently reproducible outcomes, including review quality and rework, rather than pull-request volume alone.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Claude](https://newruntime.com/topics/claude.md): Explore the claude topic hub.
- topic hub: [Claude Code](https://newruntime.com/topics/claude-code.md): Explore the claude-code topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)