ar-closeout

Summarize perf autoresearch results, verify final state, and archive artifacts for handoff.

Updated Aug 27, 2026

Search Tags:#workflow #verification #handoff #archive #ar-closeout #perf-autoresearch

Ready to run

npx skills add https://github.com/Elijah-J/InfoScraper --skill ar-closeout

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ar-closeout
Source: https://github.com/Elijah-J/InfoScraper/tree/main/.agents/skills/ar-closeout
Command: npx skills add https://github.com/Elijah-J/InfoScraper --skill ar-closeout

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Close out the completed perf autoresearch project by summarizing results, verifying state, archiving artifacts, and preparing a handoff for downstream tasks.

Core Features & Use Cases

Summarize results, best keep stacks, and exit reasons to preserve the benchmark record.
Verify finished benchmark state by re-running the final eval and confirming alignment with claimed bottlenecks.
Archive relevant artifacts (results.tsv, last_eval.md/json, and corpus notes) and prepare a structured handoff to ar-archive.
Hand off to the ar-archive stage and, if requested, prepare the commit/merge/push sequence for deployment.

Quick Start

Run ar-closeout on the completed autoresearch project directory to generate the archive bundle and handoff artifacts.

Frequently Asked Questions about ar-closeout

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I close out a completed perf autoresearch project?▼

To close out a completed perf autoresearch project, you summarize results, verify the final benchmark state by re-running the last evaluation, archive artifacts like results.tsv, and prepare a structured handoff for downstream archival processes.

What is included in an autoresearch archive bundle?▼

An autoresearch archive bundle includes results.tsv, last_eval files in markdown or json formats, and corpus notes, ensuring the benchmark record and best keep stacks are preserved for future reference.

How do I verify the final benchmark state before archiving autoresearch results?▼

You verify the final benchmark state by re-running the final evaluation and confirming that the output aligns with the claimed bottlenecks and exit reasons recorded during the autoresearch loop.

Can I automate the commit and merge sequence when wrapping up an autoresearch loop?▼

Yes, you can prepare an optional commit, merge, and push sequence for deployment when wrapping up an autoresearch loop, provided the final state verification and archive handoff are complete.

What guardrails exist to preserve the benchmark record during autoresearch handoff?▼

Guardrails during autoresearch handoff ensure the benchmark record is preserved by structuring the archive bundle and passing it safely to the ar-archive stage without altering the original evaluation data.