agentsop-conventions-pinning

Write and pin project-level conventions files for multi-tool coder-agent environments.

287|16|Updated May 20, 2026
One-click install
npx skills add https://github.com/agentsope/SkillAlchemy --skill agentsop-conventions-pinning
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agentsop-conventions-pinning
Source: https://github.com/agentsope/SkillAlchemy/tree/main/skills/agentsop-conventions-pinning
Command: npx skills add https://github.com/agentsope/SkillAlchemy --skill agentsop-conventions-pinning

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Prevents coder-agents from silently ignoring your project’s coding standards by turning your conventions into a reliably loaded, tool-compatible, conflict-aware guidance file.

Core Features & Use Cases

  • Conventions-file blueprint: defines what belongs in CONVENTIONS.md/CLAUDE.md/.cursor rules/.clinerules/AGENTS.md and what should be omitted to avoid context bloat.
  • Multi-tool loading SOP: covers the four common persistence mechanics—explicit read-only attachment, Claude Code ancestor-walk auto-load, glob-scoped rules, and CrewAI backstory injection.
  • Evidence-based evolution loop: provides an A/B verification approach plus when to add, delete, or split rules to reduce drift and contradictions over time.
  • Conflict resolution guidance: specifies how to handle precedence when pinned conventions contradict existing repository patterns, preventing arbitrary tool behavior.

Quick Start

Ask an agent to write and load your conventions file by following the conventions-pinning skill so the agent passes a targeted A/B test for a single rule (for example, using httpx instead of requests).

Frequently Asked Questions about agentsop-conventions-pinning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make coder-agents follow my project's style guide across repeated sessions?

To make coder-agents follow your project's style guide, you write and pin a conventions file using a tool-agnostic SOP, ensuring reliable loading across repeated sessions. This prevents agents from silently ignoring your coding standards.

What is the best way to maintain multi-tool interoperability for conventions files in AI coding?

Multi-tool interoperability for conventions files is maintained by using a blueprint that defines what belongs in CONVENTIONS.md or .cursor rules, satisfying tool-specific loading requirements and conflict precedence without causing context bloat.

How do I load a conventions file in Claude Code, Cursor, and CrewAI simultaneously?

You load a conventions file across these tools by following a multi-tool SOP that covers explicit read-only attachment, Claude Code ancestor-walk auto-load, glob-scoped rules, and CrewAI backstory injection mechanics for consistent guidance.

Does writing a CONVENTIONS.md file work with Aider and Cline for context engineering?

Yes, writing a CONVENTIONS.md file works with Aider and Cline. A tool-agnostic SOP ensures these coder-agents load a consistent conventions source for guidance, satisfying their specific file persistence and loading requirements.

Why does my coder-agent ignore existing repository patterns when applying new conventions?

Your coder-agent ignores existing repository patterns because of unspecified conflict precedence. You need conflict resolution guidance that specifies how to handle precedence when pinned conventions contradict existing patterns, preventing arbitrary tool behavior.

How to verify if a coder-agent is actually applying a specific coding rule from the context file?

To verify if a coder-agent applies a specific coding rule, you run a concrete A/B behavior test with controlled rule specificity. This evidence-based evolution loop checks whether the agent passes a targeted test, like using httpx instead of requests.