agent-md-refactor
Refactor Agent Instruction Files
Large agent instruction files waste context and hide important rules. This skill splits them into a compact root file and focused linked documents.
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Review the Skillstore skill "agent-md-refactor" from https://skillstore.io/skills/softaworks-agent-md-refactor.md and its manifest at https://skillstore.io/api/skills/softaworks-agent-md-refactor/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "agent-md-refactor". A 300-line CLAUDE.md mixes project overview, testing rules, style rules, and workflow notes.
Expected outcome:
- A short root file keeps the project summary and essential commands.
- Linked topic files hold testing, style, architecture, and workflow guidance.
- A deletion report lists vague or redundant instructions.
Using "agent-md-refactor". An AGENTS.md file contains conflicting style rules and repeated testing guidance.
Expected outcome:
- The skill reports the conflicting rules and asks which one should win.
- After resolution, it groups the remaining guidance into clear topic files.
- The final checklist verifies links, coverage, and root file size.
Using "agent-md-refactor". A team wants one structure for agent instructions across several repositories.
Expected outcome:
- Each repository receives the same root file pattern and topic categories.
- Project-specific exceptions stay in the relevant linked files.
- Teams can compare guidance without reading one large file.
Security Audit
SafeI found no actionable security issue in the static findings. All Ruby, shell, and reconnaissance alerts point to Markdown examples, table headings, or documentation text. No prompt injection, data exfiltration, or unauthorized execution intent was found in README.md or SKILL.md.
Risk Factors
⚙️ External commands (36)
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/softaworks-agent-md-refactor/audits/4?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/softaworks-agent-md-refactor?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/softaworks-agent-md-refactor?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/softaworks-agent-md-refactor/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/softaworks-agent-md-refactor.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
softaworks. (2026). agent-md-refactor security audit report (audit version 4) [Author version unspecified]. Skillstore. https://skillstore.io/skills/softaworks-agent-md-refactor/audits/4BibTeX citation
@techreport{softaworks-softaworks-agent-md-refactor-2026,
author = {softaworks},
title = {agent-md-refactor security audit report (audit version 4)},
institution = {Skillstore},
year = {2026},
number = {4},
url = {https://skillstore.io/skills/softaworks-agent-md-refactor/audits/4},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "agent-md-refactor security audit report (audit version 4)"
version: "unspecified"
type: report
authors:
- name: "softaworks"
date-released: "2026-07-07"
url: "https://skillstore.io/skills/softaworks-agent-md-refactor/audits/4"
identifiers:
- type: other
value: "skillstore:softaworks-agent-md-refactor:audit:4"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: HighWhat You Can Build
Clean Up CLAUDE.md
Turn a long CLAUDE.md file into a short root guide with linked topic files.
Organize AGENTS.md
Group mixed coding, testing, style, and workflow rules into maintainable documents.
Standardize Team Guidance
Create a consistent structure for agent instructions across multiple repositories.
Try These Prompts
Refactor my CLAUDE.md using progressive disclosure. Keep only universal guidance in the root file.
Review my AGENTS.md and split it into focused linked files for testing, style, architecture, and workflow.
Analyze my agent instructions for contradictions first. Ask me to resolve conflicts before you restructure the files.
Audit all agent instruction files in this repository. Propose a minimal root file, linked topics, deletion candidates, and verification checks.
Best Practices
- Start from a clean git state so changes are easy to review.
- Resolve contradictions before moving content into linked files.
- Keep the root file limited to guidance needed for every task.
Avoid
- Keeping every project rule in the root instruction file.
- Creating many small files that make guidance harder to find.
- Preserving vague instructions that do not change agent behavior.
Frequently Asked Questions
What files does this skill help refactor?
Does it delete instructions automatically?
Can it handle conflicting rules?
What should stay in the root file?
How many linked files should I create?
Does it work with Claude, Codex, and Claude Code?
Developer Details
Author
softaworksLicense
MIT
Skillstore revision
r1
Version notice
The author did not declare a version.
Ref
f93e9bb0daca99badb6a7e574b97737155d57cb3
Maintenance freshness
7/18/2026
Usage
14 downloads · 122 views