Skip to content
Agentmart
Submit a tool

refine

Analyze past cueloop Threads and propose recurring feedback writebacks through a plan Thread.

Skill by mmurakaru

IndexedApache-2.0★ 8
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/mmurakaru-refine

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Analyze past cueloop Threads and propose recurring feedback writebacks through a plan Thread. Use when the user asks to learn from past reviews or run refine.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/refine

What's inside

Skill name
refine
Description
Analyze past cueloop Threads and propose recurring feedback writebacks through a plan Thread. Use when the user asks to learn from past reviews or run refine.
License
Apache-2.0
Instructions
21 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# cueloop refine

Use the harness's cueloop `refine` workflow when available. Otherwise, run
`cueloop refine` and read the report path in its JSON output to analyze the
stored Thread corpus.
Read its report and group annotations into named patterns. A pattern needs at
least three members; keep one-offs in a long-tail list. Rank patterns by how
often they occur on changes-requested Messages.

Draft one proposed writeback per pattern. Include its evidence, target file,
and exact text. Route repo-specific lessons to that repo's `AGENTS.md` or a
project skill; route global preferences to the appropriate global instruction
or memory. Submit the proposals as a plan Thread through the same workflow,
or write them to a file and run
`cueloop session create --type plan --content-file <proposals.md>`. On the CLI
path, collect the Message with `cueloop session wait <id> --timeout-ms 60000`.

Do not edit the targets before approval. Corpus analysis does not mutate source
Threads; only approved writebacks change instructions or memory.

From mmurakaru/cueloop, Apache-2.0.

Signals

Stars
8
Forks
0
Open issues
16
Last commit (this folder)
Sep 21, 2026 (1 month ago)
Last release
Oct 20, 2026 (9 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
Apache-2.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from mmurakaru

Cite this page

refine skill. Agent Mart. https://agentmart.dev/skills/mmurakaru-refine (last updated 2026-10-24)