From devflow
Search MemPalace for project memories — agent diaries, PREVC decisions, and historical context
npx claudepluginhub nexuz-sys/devflow --plugin devflowThis skill uses the workspace's default tool permissions.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.
Executes pre-written implementation plans: critically reviews, follows bite-sized steps exactly, runs verifications, tracks progress with checkpoints, uses git worktrees, stops on blockers.
On-demand deep-dive into MemPalace memories. Use when auto-recall budget is insufficient or when you need to explore historical context.
Announce at start: "I'm using the devflow:memory-recall skill to search project memories."
Check if MemPalace MCP is available:
mempalace: true in session context → proceed/devflow init to set it up."--global flag)Accept natural language queries:
Read mempalace config from .context/.devflow.yaml:
mempalace:
palace: ~/.mempalace/palace # palace path
wing: auto # project wing (auto = repo name)
Default: search within project wing only.
If user passes --global or asks for cross-project context, search all wings.
Call MemPalace MCP search tool:
Group results by type:
[2026-04-08 14:30] architect (Planning)
Summary: Defined hybrid MCP-first architecture for mempalace integration
Decisions: MCP-first without CLI fallback, 500 token budget for auto-recall
Artifacts: docs/superpowers/specs/2026-04-08-mempalace-design.md
→ Handed off to: code-reviewer
[2026-04-08 15:00] Review phase — mempalace-integration
Approved: MCP-first approach, no Python dependency
Flagged: Consider token budget monitoring
Chronological view when user asks for "history" or "timeline".
After presenting results, offer refinement: