SKILLEMALL.ai
GitHub ↗ RU Sign in

CB loopx

Adopt and operate LoopX for durable goals, long-running agent work, scoped collaboration, evidence-backed progress, and recovery across sessions. Use when the user asks to set up LoopX, start or resume LoopX work, inspect progress, or select a LoopX workflow.

huangruiteng/loopx Agent Skills author: huangruiteng Apache-2.0 34 files · 3 scripts body ≈ 5 404 tokens Open the sourcegithub.com↗ analyzed 6 h ago

Adopt and operate LoopX for durable goals, long-running agent work, scoped collaboration, evidence-backed progress, and recovery across sessions.

As a process B 66/100 · Nearly there — weak spots: result and completion, consistency, running it twice

ProcedureGitHubAI and agentstype and topics are labelled automatically from the skill text
Runs in: Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) OpenClaw Goose OpenCode
JSON
Technical rating
C
80/100
safety, quality, tests
Safety 60%
84
Quality 40%
74
Run on models
none yet
Process rating
B
66/100
Nearly there
Progress reporting w 2
0
Running it twice w 4
30
Result and completion w 14
40
the three weakest of ten parameters · all ten

How to improve

  1. The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
For the model run — optional
  • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
  • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

Guard findings · 16

✓ No critical or high findings

Medium and low: 16
  • low Obfuscation obf-base64-blob blog/application-scenarios/index.html:235
    Long base64-looking blob (detector / deny-list definition)
    <li><a href="../../benchmarks/lhtb/?lang=en">LHTB: compared with Plain and native Goal</a>; <a href="https://github.com/huangruiteng/loopx/blob/d608…c40/benchmark/LHTB/
    detector
  • low Obfuscation obf-base64-blob blog/application-scenarios/index.html:239
    Long base64-looking blob (quoted — discussed, not commanded)
    <li><a href="https://github.com/huangruiteng/loopx/blob/ed1a…100/loopx/capabilities/explore/README.md">Explore: evidence graphs, planning, and authority boundaries</a>.
    quoted
  • low Obfuscation obf-base64-blob blog/from-one-shot-agents-to-long-horizon-control/index.html:193
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>For open-ended research, the <a href="https://github.com/huangruiteng/loopx/blob/41a3…a4b/loopx/capabilities/explore/README.md">Explore capability</a> adds an eviden
    quoted
  • low Obfuscation obf-base64-blob blog/semantic-software-development/index.html:89
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>In local LoopX development, agents use CLIs to read work, invoke tools, submit results and reconcile state, following the current <code>interaction_contract</code>. The <a href="https://github.com/
    quoted
  • low Obfuscation obf-base64-blob blog/semantic-software-development/index.html:90
    Long base64-looking blob (detector / deny-list definition; security demo / example)
    <p>Optional <a href="https://github.com/loopx-project/loopx/blob/f713…2ba/loopx/capabilities/explore/README.md">Explore</a>, for example, organizes questions, experimen
    detectordemo
  • low Obfuscation obf-base64-blob blog/semantic-software-development/index.html:97
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>LoopX development follows a <a href="https://github.com/loopx-project/loopx/blob/f713…2ba/docs/development/README.md#change-loop--变更闭环">change loop</a> from problem
    quoted
  • low Obfuscation obf-base64-blob blog/semantic-software-development/index.html:105
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>The <a href="https://github.com/loopx-project/loopx/blob/f713…2ba/benchmark/LHTB/studies/five…max/README.md">five-arm LHTB study</a> checks whole-system
    quoted
  • low Obfuscation obf-base64-blob blog/zh/application-scenarios/index.html:232
    Long base64-looking blob (detector / deny-list definition)
    <li><a href="../../../benchmarks/lhtb/?lang=zh">LHTB：与 Plain、原生 Goal 的对比</a>；<a href="https://github.com/huangruiteng/loopx/blob/d608…c40/benchmark/LHTB/studies/five-ar
    detector
  • low Obfuscation obf-base64-blob blog/zh/application-scenarios/index.html:236
    Long base64-looking blob (quoted — discussed, not commanded)
    <li><a href="https://github.com/huangruiteng/loopx/blob/ed1a…100/loopx/capabilities/explore/README.md">Explore：证据图、规划与权限边界</a>。</li>
    quoted
  • low Obfuscation obf-base64-blob blog/zh/from-one-shot-agents-to-long-horizon-control/index.html:195
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>对开放式研究，<a href="https://github.com/huangruiteng/loopx/blob/41a3…a4b/loopx/capabilities/explore/README.md">Explore Capability</a>补充问题、假设、实验和 Finding 的证据图。规划层可以组织分支，执行
    quoted
  • low Obfuscation obf-base64-blob blog/zh/semantic-software-development/index.html:89
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>本机推进 LoopX 时，Agent 通过 CLI 读取工作、调用工具、提交结果和回读状态；下一步采用当次 <code>interaction_contract</code>。<a href="https://github.com/loopx-project/loopx/blob/f713…2ba/loopx/capabilit
    quoted
  • low Obfuscation obf-base64-blob blog/zh/semantic-software-development/index.html:90
    Long base64-looking blob (security demo / example; quoted — discussed, not commanded)
    <p>例如可选的 <a href="https://github.com/loopx-project/loopx/blob/f713…2ba/loopx/capabilities/explore/README.zh-CN.md">Explore</a>把问题、实验、发现和反证组织成证据图；PR Review 则组织具体版本的审阅计划、
    demoquoted
  • low Obfuscation obf-base64-blob blog/zh/semantic-software-development/index.html:97
    Long base64-looking blob (quoted — discussed, not commanded)
    <p>LoopX 自身的<a href="https://github.com/loopx-project/loopx/blob/f713…2ba/docs/development/README.md#change-loop--变更闭环">变更闭环</a>从问题刻画走向聚焦实现、回归检查、按风险选择 canary 和审阅。<code>
    quoted
  • low Obfuscation obf-base64-blob blog/zh/semantic-software-development/index.html:105
    Long base64-looking blob (quoted — discussed, not commanded)
    <p><a href="https://github.com/loopx-project/loopx/blob/f713…2ba/benchmark/LHTB/studies/five…max/README.md">LHTB 五臂研究</a>提供了一种跨版本、跨执行方式的整体检查：46 个共同任务覆盖 Pla
    quoted
  • low Secrets in code secret-high-entropy-token lab/time-machine/data/graph.json:2930
    High-entropy token-like string (may be an id, hash or a credential) (detector / deny-list definition)
    "summary": "Input revision: 8352…4bc Applicability: Immutable53 8352…c08 versus10;36small/16actual weighted references,8fresh5201/2089prior
    detector
  • low Secrets in code secret-high-entropy-token lab/time-machine/data/graph.json:3029
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "summary": "Input revision: a7e7…629 Applicability: Immutable54 a7e7…629;12owned5201/208
    quoted

Files scanned: 34. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning body-long SKILL.md body ≈ 5404 tokens (recommended < 5000); move details to references/

Process rating: all ten parameters 66/100

  • 0Progress reporting. Says nothing while it works
  • 30Running it twice. 7 mutating operations with no state check
  • 40Result and completion. Does not say what the result is
  • 40Consistency. Frontmatter name (loopx) differs from the folder (public)
  • 60Tools and files. Uses tools (bash, web, python) that frontmatter does not declare
  • 70When it triggers. States when to use, but not when not to
  • 70Inputs and preconditions. Inputs and preconditions are listed
  • 70Execution cost. Instruction body is 5404 tokens
  • 100Steps. 30 steps
  • 100Failures and branches. 1 branches, has a failure section
  • low 12 top-level sections: this looks like several domains in one skill
  • high The skill tells the model to perform an irreversible action with no human approval

Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

Quality signals

  • +5Description has no quoted example phrases that should trigger the skill
  • +4Description does not say when NOT to use the skill (false activations)
  • +3Output format is not stated: the model decides each time
  • +1No license
  • +2Single-language instructions
  • +3Description length 259: enough signal without eating the budget
  • +4Structure: 14 headings
  • +3Step-by-step instructions: 30 items
  • +4Has examples (6 code blocks)

Quality base 70; lint remarks subtract, signals add up to 100. Result: 74.