Skills Overview
All skills available in Coding Friend including slash commands and auto-invoked skills.
Coding Friend provides a comprehensive toolkit of 19 skills that automate common engineering workflows. Sixteen are available as slash commands (some also auto-invoke), while three activate only automatically in response to your coding patterns.
Slash Commands
Interactive skills you trigger directly. Some also auto-invoke — they activate automatically when relevant context is detected, without needing you to type the command.
| Command | Description | Triggered by | Context |
|---|---|---|---|
| /cf-ask | Quick Q&A about your codebase with persistent memory | slash | Medium~1K–2.5K tokens injected into prompt |
| /cf-commit | Smart conventional commits with diff analysis | slash | Medium~1K–2.5K tokens injected into prompt |
| /cf-fix | Quick bug fix workflow with problem verification | slash + auto | High>2.5K tokens injected into prompt |
| /cf-help | Answer questions about Coding Friend | slash + auto | Medium~1K–2.5K tokens injected into prompt |
| /cf-learn | Extract human learning docs from coding sessions | slash + auto | Medium~1K–2.5K tokens injected into prompt |
| /cf-optimize | Structured optimization with before/after measurement | slash + auto | Medium~1K–2.5K tokens injected into prompt |
| /cf-plan | Brainstorm and create structured implementation plans | slash + auto | High>2.5K tokens injected into prompt |
| /cf-remember | Capture project knowledge for AI memory across sessions | slash + auto | High>2.5K tokens injected into prompt |
| /cf-research | In-depth research with web search and parallel subagents | slash | Medium~1K–2.5K tokens injected into prompt |
| /cf-review | Multi-layer code review in a forked subagent | slash + auto | Medium~1K–2.5K tokens injected into prompt |
| /cf-review-in | Collect external review results | slash | Medium~1K–2.5K tokens injected into prompt |
| /cf-review-out | Generate review prompt for external AI | slash | Low<1K tokens injected into prompt |
| /cf-scan | Scan project and bootstrap memory with project knowledge | slash | High>2.5K tokens injected into prompt |
| /cf-session | Save session to docs/sessions/ to resume on another machine | slash | Medium~1K–2.5K tokens injected into prompt |
| /cf-ship | Verify, commit, push, and create PR in one command | slash | Low<1K tokens injected into prompt |
| /cf-teach | Personal teacher — conversational storytelling breakdown of any task | slash | Medium~1K–2.5K tokens injected into prompt |
| /cf-warm | Catch up after absence — summarize git history for a user | slash | High>2.5K tokens injected into prompt |
Auto-Invoked Skills
These skills activate automatically and cannot be triggered manually. They have no slash command.
| Skill | Activates When | What It Does | Context |
|---|---|---|---|
| cf-sys-debug | Debugging issues | Guides 4-phase systematic debugging + documentation: Investigate, Analyze, Test, Fix, then Document | Medium~1K–2.5K tokens injected into prompt |
| cf-tdd | Writing new code | Enforces test-driven development: RED → GREEN → REFACTOR | Medium~1K–2.5K tokens injected into prompt |
| cf-verification | Before claiming task complete | Ensures tests pass, changes verified, no regressions introduced | Low<1K tokens injected into prompt |
Context column shows the approximate prompt footprint (tokens injected when loaded): ⚡ = low, ⚡⚡ = medium, ⚡⚡⚡ = high. See Context Footprint for details.
Custom Skill Guides
You can extend any built-in skill with your own guidance by creating a <skill-name>-custom/ directory with a SKILL.md file in .coding-friend/skills/ (per-project) or ~/.coding-friend/skills/ (global). Each guide supports 3 optional sections:
## Before— runs before the builtin workflow starts## Rules— additional rules applied throughout the workflow## After— runs after the final step completes
For example, .coding-friend/skills/cf-commit-custom/SKILL.md can add custom commit rules or trigger tests after each commit. See Custom Skill Guides for full details.