/cf-remember
High>2.5K tokens injected into promptCapture project knowledge for AI memory across sessions.
Context footprint: ⚡⚡ (medium) — what does this mean?
The /cf-remember skill captures important project knowledge and saves it for AI recall in future sessions in docs/memory/. Unlike /cf-learn (educational notes for humans), this saves project context — conventions, decisions, feature flows, gotchas — that Coding Friend and Claude Code can reference later.
Usage
/cf-remember [topic]
Without [topic], it will scan the entire conversation for key knowledge.
What It Does
- Captures Context — Extracts relevant code patterns and decisions
- Documents Learning — Writes clear explanations of how things work
- Saves to Memory — Stores in
docs/memory/with proper organization - Indexes for Retrieval — Makes knowledge searchable and linkable
- Enables Cross-Session Reference — Future work uses these docs
Examples
/cf-remember auth flow decisions and implementation
/cf-remember
# Auto detect the key knowledge based on the conversation
Memory Organization
Knowledge is saved to docs/memory/ organized by category:
| Category | Folder | Use For |
|---|---|---|
| Features | features/ | Feature-specific logic, flows, APIs |
| Conventions | conventions/ | Project-wide patterns and rules |
| Decisions | decisions/ | Architecture/design decision records |
| Bugs | bugs/ | Bug root causes, fixes, and how to avoid recurrence |
| Infrastructure | infrastructure/ | Build, deploy, CI/CD, environment setup |
Category selection guide:
- Fixing a bug →
bugs/(root cause, fix, prevention), NOTfeatures/ - Adding/changing a feature →
features/ - Establishing a pattern or rule →
conventions/ - Choosing between approaches →
decisions/ - Build/deploy/tooling knowledge →
infrastructure/
For details on memory types, markdown format, search pipeline, and tier architecture, see the Memory System reference.
Auto-Invocation
cf-remember also activates automatically when the conversation produces significant project knowledge worth preserving — architecture decisions, non-obvious conventions, complex feature flows, or important gotchas that would save time in future sessions. It does NOT auto-invoke for trivial changes.
When to Use
- After onboarding to understand current patterns
- When completing complex features
- After discovering important design decisions
- When solving challenging problems (for future reference)
- Documenting gotchas and lessons learned