new-loop
Spin up a new loop (domain) in a file-based knowledge base — bootstrap the substrate if it's missing, gather the loop's charter, scaffold domains/<loop>/README.md, then do ONE real test run and record it in the loop's Timeline and LOG.md
Install / Use
npx skills add AI-Builder-Club/skills --skill new-loopInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
Development & EngineeringSupported Platforms
Tags
Our assessment of new-loop
new-loop scores 82/100 on our quality scale, 3377th of 4,700 Development & Engineering skills we index.
Its SKILL.md is 4.8 KB long, well organised into 9 sections and no code examples: a solid amount of guidance for an agent.
With 1,274 GitHub stars, it is one of the more widely adopted skills in the catalogue.
Maintenance, license and trust
- The repository was last updated about 2 months ago, so new-loop is actively maintained.
- Our last check on 2026-11-07 found the source still online.
- No license is declared. By default that means all rights are reserved: you can read it, but reusing or redistributing it is not clearly permitted. Ask the author before building on it commercially.
- Its trust signals score 88/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-10-03. Automated pattern scan on 2026-10-03. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
new-loop compared with similar skills
All 4 of these similar skills score higher than new-loop; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| new-loop (this skill)by AI-Builder-Club | 82 | 1.3k | 2mo ago | SKILL.md |
| taste-skillby Leonxlnx | 100 | 116.8k | 8d ago | CLAUDE.md |
| ai-job-searchby ai-job-search-dev | 100 | 72.1k | today | CLAUDE.md |
| claude-howtoby luongnv89 | 100 | 47.3k | 2d ago | CLAUDE.md |
| lobehubby lobehub | 100 | 85.6k | today | MCP Server |
Frequently asked questions
- How do I install new-loop?
- Run
npx skills add AI-Builder-Club/skills --skill new-loop. The install tabs above show the steps for each supported agent. - Which AI agents does new-loop work with?
- It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is new-loop safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It declares no license and scores 88/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is new-loop still maintained?
- The repository was last updated about 2 months ago, so new-loop is actively maintained.
Skill content
View source on GitHubname: new-loop description: Spin up a new loop (domain) in a file-based knowledge base — bootstrap the substrate if it's missing, gather the loop's charter, scaffold domains/<loop>/README.md, then do ONE real test run and record it in the loop's Timeline and LOG.md. Use when the user says "set up a new loop", "create a domain", "start a new beat/workstream", or names a recurring job they want the agent to own. user_invocable: true
new-loop — spin up a new loop
A loop (a domain) is a recurring thread of work the agent owns: a charter, a cadence, and
the artifacts it produces. This skill creates one, proves it works with a single real run, and
leaves behind a domains/<loop>/README.md that is the loop's live state.
When to use
The user wants to stand up a new workstream/beat/job (e.g. "a weekly SEO loop", "a support
triage loop", "a competitor-watch loop"). Don't use this for a one-off task — that's a backlog
line in an existing domain, or a doc/signal.
Inputs to gather (ask only what's missing)
Infer from the request; ask a short clarifying round only for what you can't:
- name — kebab-case, the loop's home folder (
domains/<name>/). Keep it short. - goal — one line: the outcome this loop drives.
- cadence —
manual/daily/weekly/ a cron expr. Defaultmanual. - what it does — what it consumes (signals? data? an inbox? a URL?) and produces (signals?
docs? a report? code changes shipped via
/verify?). - tools/data — sources or credentials it needs (point at a setup skill or
.env; never inline secrets).
If the request is already specific, infer all five and just confirm in your summary.
Procedure
1. Bootstrap the substrate (one-time; skip if already set up)
Check the knowledge-base repo root for:
ARCHITECTURE.mdandLOG.md, and- a
CLAUDE.mdthat has a "Knowledge base" section.
All present → the substrate exists; skip to Step 2.
Anything missing → read references/KNOWLEDGE_SETUP.md and follow it — it copies in
ARCHITECTURE.md + LOG.md, creates signals/ docs/ domains/ with their README schemas, and
injects the knowledge-base section into CLAUDE.md (or scaffolds one from
references/CLAUDE.template.md). It's idempotent: it only creates what's missing.
(Read references/ARCHITECTURE.md once if you haven't — it's the model this skill instantiates.)
2. Scaffold the loop README
Create domains/<name>/README.md from the domain template (in domains/README.md, also
quoted in references/KNOWLEDGE_SETUP.md), filled with the gathered inputs. Required sections:
frontmatter (kind: domain, domain, status: active, goal, cadence), a 2–4 line
description, ## Current focus, ## Backlog (to-dos inline — they stay in the README until they
earn a task kind), and an empty ## Timeline. Add ## Evidence & analysis / ## Metrics
placeholders if relevant.
Collision check: if domains/<name>/ already exists, stop and ask whether to update it instead
of overwriting.
3. Do ONE real test run
The point of the skill: prove the loop actually runs, not just that the folder exists.
Actually run the loop once, at small scale — do whatever it's meant to do (triage a few real tickets, pull one real SERP, fetch the inbox, draft one comment, run one analysis query, scope one code change…). Use real tools/data where you can; if a credential is missing, do the furthest-reachable dry run and note the gap.
Producing an artifact is optional — a legit run may surface nothing worth filing. Only create
a signal/doc if the run genuinely produced one.
Two required outputs regardless:
- Append one dated line to the loop README's
## Timeline:YYYY-MM-DD | test run — <what you did and found / "nothing actionable yet">. - Append one entry to
LOG.md(its grammar):## YYYY-MM-DD · <loop-name> loop created + first run · #ops What: <one line — what the loop is and what the first run did/found>. Refs: domains/<name>/README.md (new)[, any artifact created].
4. Report back
Summarize: the loop's charter (the five inputs), what the test run did/found, any artifacts created (or "none — nothing actionable this run"), missing tools/credentials to wire up, and how to run it again (cadence + entry point). Keep it tight.
Notes
- Don't gold-plate the scaffold. A loop README is live state, not a spec — start lean; let it accrete via its Timeline.
- One loop = one separable workstream. If what the user described is really part of an
existing loop, add it there (a backlog line + a
domain:tag) instead of a near-duplicate. - For loops that ship code, the loop's run works in an isolated git worktree and ships via
the
/verifyskill (proof + PR), giving each parallel agent its own isolated stack viacrabbox-setup(sibling harness skills in this plugin). Point the README's Backlog at them.
Related Skills
taste-skill
116.8kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
ai-job-search
72.1kThe job search that runs on your machine. AI job application framework built on Claude Code: evaluate postings, tailor CVs, write cover letters, prep interviews. Fork it and own it.
claude-howto
47.3kA visual, example-driven guide to Claude Code — from basic concepts to advanced agents, with copy-paste templates that bring immediate value.
lobehub
85.6k🤯 LobeHub - Super-Agents on Your Team. Add super-agents to the chats your team already uses.They take on work by themselves and get better every week.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
