Recall Feature Context

Load context from previous work sessions so you can pick up where you left off without grepping the entire codebase.

Context Storage

The memory root is this session's auto-memory directory — the path named in your system prompt's memory section, i.e. ~/.claude/projects/<slug>/memory/, where <slug> is the absolute working directory with every non-alphanumeric character replaced by -.

Feature context files live in <memory root>/features/, one file per feature named {feature-name}.md (kebab-case). If that directory doesn't exist, no context has been saved for this project yet — say so and proceed with normal codebase exploration.

When This Skill Is Invoked

1. List available context files in <memory root>/features/ using Glob
2. Match the relevant file based on what the user is asking about (the feature name, the branch name, or keywords from their request)
3. Read the matching context file(s) to load prior session knowledge
4. Check recent git history on the current branch for additional context: git log --oneline -20
5. Summarize what you know to the user before starting work — confirm you have the right context

If no context file matches, tell the user and proceed with normal codebase exploration.

After Completing Significant Work

When this skill is active and you complete significant feature work in a session, you MUST update or create the feature context file before finishing. This is non-negotiable. Follow the remember skill — it owns the template and the rules for what belongs in a context file.

Briefly: write the feature name in kebab-case, follow <memory root>/features/_template.md, merge into an existing file rather than overwriting it, and record decisions, corrections, key file paths, deferred work, patterns, and gotchas. The file is a dated decision record, not a status report — don't save progress state, framework conventions, file contents, or in-flight PR status.