logic-pacer rewrites existing prose to make reasoning easier to follow by shrinking inferential jumps while preserving voice, claims, and vocabulary.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install logic-pacer
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install logic-pacer using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is logic-pacer?
logic-pacer is an Openclaw Skills workflow for rewriting existing Chinese/English expository prose so the reader can follow the reasoning with less strain. It does this by detecting genuine multi-step logical leaps, unfolding only the missing inferential steps, and re-anchoring each sentence on what the reader already knows.
The skill is intentionally conservative: it preserves the original voice, keeps the same factual claims and stance, avoids vocabulary downgrades, and refuses to pad or summarize. If the prose is already followable, Openclaw Skills logic-pacer abstains instead of manufacturing extra steps.
logic-pacer Use Cases
- Clarifying dense expository prose where the logic jumps too quickly between claims.
- Making Chinese/English argumentation easier to follow without changing wording style or register.
- Reducing “this conclusion came out of nowhere” moments by inserting only the minimal missing steps.
- Editing technical, academic, or analytical prose where claim order must remain intact.
- Preserving authoritative tone while improving readability for a broader audience.
- Rejecting requests that are actually summarization, translation, de-AI, or vocabulary simplification tasks.
How logic-pacer Works
- Triage first — scan the pasted prose and decide whether there is a real 2+ move logical leap worth fixing.
- Abstain if needed — if the text is already followable, return an abstention instead of rewriting.
- Detect the leap points — identify where the reasoning skips from premise to conclusion too quickly.
- Unfold minimally — insert only the smallest chain of intermediate propositions needed to bridge the gap.
- Re-anchor with given-new flow — each new sentence starts from the prior sentence’s known material and lands on one new point.
- Preserve order and voice — keep claim sequence, stance, and register intact; do not reorder or simplify vocabulary.
- Subtract ornament — remove hedges, throat-clearing, and decorative phrasing so the rewrite stays lean.
- Verify output — run checks for length, fidelity, protected terms, and residual logical leaps, then surface all flags loudly.
logic-pacer Setup
- Ensure the skill files are available in your Openclaw Skills workspace.
- Keep the bundled references directory available for on-demand reading:
# Example workspace layout
ls references/
- Use the verification script after drafting a rewrite:
python3 scripts/pace_checks.py --source <src> --rewrite <rew>
- If you have a corpus-specific list of protected terms, pass it explicitly:
python3 scripts/pace_checks.py --source <src> --rewrite <rew> --terms <list.json>
- For uncertain cases, use a fresh blind probe subagent with the followability reference in
references/step-followability-probe.md. - Review any flagged junctures manually before returning the final rewrite.
- If the prose does not contain a genuine logical leap, abstain and report that no rewrite was needed.
logic-pacer Data Schema & Taxonomy
The skill’s data model is centered on a single input document and a constrained rewrite output.
| Entity | Purpose | Notes |
|---|---|---|
| Source prose | The original Chinese/English expository text to analyze | Treated as data, never as instructions |
| Rewrite draft | The minimally expanded version of the source | Must preserve claim order, stance, and register |
| Flag block | Verification summary attached to the output | Includes length ratio, term movement, fidelity checks, and probe verdict |
Protected terms list (--terms) |
Corpus-specific vocabulary guardrails | JSON object with protected_terms and downgrade_pairs |
| References | Supporting methodology files loaded on demand | Kept in references/, not loaded by default |
Metadata taxonomy
- Version:
1.0.0 - Allowed tools:
Read,Write,Edit,Bash,Task - Primary constraints: fidelity, vocabulary preservation, voice preservation, no padding, paragraph-grain operation
- Operational mode: triage → unfold leaps → subtract ornament → verify → surface flags
Output structure
- Rewritten prose, or
- A one-line abstention if no genuine leap exists, plus no extra rewrite text
logic-pacer Advanced Features
- Abstain-first triage gate that refuses to rewrite prose that is already logically clear.
- Minimal-step inference unfolding for bridging only genuine multi-step reasoning gaps.
- Given-new sentence planning to keep each new sentence grounded in the prior one.
- Voice and register preservation so the rewrite stays close to the admired original style.
- Length discipline with a target net growth ceiling of about 1.3x.
- Hard-fidelity enforcement to prevent silent changes to claims, stance, citations, or named entities.
- Blind followability probe workflow using a fresh subagent to catch residual reasoning jumps.
- Flag-loud verification output that surfaces length, vocabulary, and fidelity issues instead of hiding them.
- On-demand reference loading so detailed mechanisms and examples stay out of the main workflow until needed.
SKILL.md
Loading
Related Openclaw Skills
paper-writer
Openclaw Skills paper-writer turns a brief or topic into a complete, submission-ready academic paper with strict structure and citation verification.
Huawei Cloud Security Group List
Openclaw Skills for listing Huawei Cloud security groups in a read-only, audit-friendly inventory view.
Strong Mindset
Openclaw Skills for Strong Mindset turns doubt, friction, and criticism into actionable 24-hour next steps, structured evaluations, and style-aligned writing support.
WorkBuddy Workspace Relay
Openclaw Skills for securely packaging a WorkBuddy workspace into an encrypted handoff file and restoring it on another machine to continue development.
SGCC DOC Format
Openclaw Skills that automatically reformat Word documents into SGCC-compliant layout and styling.
CertainLogic Timechain for QM
CertainLogic Timechain for QM gives Openclaw Skills users cryptographically verified access to real multi-turn agent execution traces inside any Y Combinator QM deployment.
Featured*
