<!-- SKILL.md is the main instruction file for a skill. It's loaded into context when the user invokes this skill. Think of skills as specialized "modes" for Claude Code — each skill gives Claude domain expertise, workflows, and guardrails for a specific task. Real-world examples: code-review, research, copywriting, deployment, SEO audit. -->
Scanned 5/27/2026
Install via CLI
openskills install lukasz-fedor/claude-memory-template# Example Skill
<!--
SKILL.md is the main instruction file for a skill.
It's loaded into context when the user invokes this skill.
Think of skills as specialized "modes" for Claude Code — each skill
gives Claude domain expertise, workflows, and guardrails for a specific task.
Real-world examples: code-review, research, copywriting, deployment, SEO audit.
-->
## Before You Start
<!--
List files Claude should read before starting work.
These give Claude the context it needs to do the job well.
Reference files stay in references/ — they're loaded on demand, not always.
-->
**Read required context:**
```
references/guidelines.md # Domain-specific guidelines
```
**Read learned patterns:**
```
LEARNED.md # What worked (and didn't) in previous sessions
```
---
## Workflow
<!--
Step-by-step process Claude follows when this skill is invoked.
Be specific — Claude follows these literally.
-->
1. Read the required context files listed above
2. Understand the user's request
3. Follow the rules below
4. Verify output against the checklist
---
## Rules
<!--
Guardrails and conventions for this skill.
These prevent common mistakes and keep output consistent.
Write them as clear imperatives — "Do X", "Never Y", "Always Z".
-->
- Always verify facts before including them in output
- Use project terminology consistently (see references/)
- Ask the user when requirements are ambiguous — don't guess
- Keep output focused and actionable
---
## Self-Learning
<!--
Skills improve over time. LEARNED.md captures what Claude discovers
during real usage — patterns that work, gotchas, user preferences.
Claude updates LEARNED.md proactively after solving non-trivial problems.
This is what makes skills get better with each session.
-->
Read `LEARNED.md` before starting. Update it when you discover:
- Patterns that work well for this skill
- Common mistakes or gotchas
- User preferences learned during use
**Consolidation (target ~50 lines):**
Before adding a new entry, check file length. If approaching 50 lines:
1. Merge duplicate/overlapping entries
2. Remove entries that haven't been reinforced
3. Keep only entries that would change behavior
No comments yet. Be the first to comment!