Skip to main content
ClaudeWave
Skill2.4k repo starsupdated today

hyperflow-cache

Hyperflow-cache manages persistent project memory stored in `.hyperflow/memory/` files, enabling users to view, search, add, edit, and prune decision logs, learnings, pitfalls, and patterns without modifying source code. Use it to document non-obvious project insights, architectural rationale, and lessons learned that would otherwise be lost, keeping them organized and searchable throughout development.

Install in Claude Code
Copy
git clone --depth 1 https://github.com/jeremylongshore/claude-code-plugins-plus-skills /tmp/hyperflow-cache && cp -r /tmp/hyperflow-cache/plugins/ai-agency/hyperflow/templates/antigravity/skills/hyperflow-cache ~/.claude/skills/hyperflow-cache
Then start a new Claude Code session; the skill loads automatically.

SKILL.md

# hyperflow-cache — memory CRUD (Antigravity single-agent)

Manage `.hyperflow/memory/` entries. **Only memory files** — never source code. Follow the `hyperflow` doctrine.

## Operations

- **view / list** — print the entries in `.hyperflow/memory/{decisions,learnings,pitfalls,patterns}.md`.
- **search `<term>`** — grep the memory files; show matching entries with their file + heading.
- **add** — append a tagged entry to the right category file (decisions / learnings / pitfalls / patterns). Use the format: `## <topic>` then `- <fact> (recorded <YYYY-MM-DD>)`.
- **edit `<entry>`** — update an existing entry in place (don't duplicate).
- **prune / clear** — remove stale or wrong entries. Confirm via AskUserQuestion before a destructive clear (binary Yes/No).

## Rules

- Scope is `.hyperflow/memory/` only. Don't record what the repo/git already captures; record the non-obvious why.