@thedotmack

mem-search queries persistent cross-session memory via three-step filtering workflow: search index by term, narrow with timeline context, then batch-fetch full details. designed to minimize token spend on memory lookups.

thedotmack-mem-search retrieves work across past sessions via a three-step filter-first workflow (search index, timeline context, batch fetch) designed to minimize token usage while surfacing relevant observations and decisions from persistent memory.

weekly-digests splits a project's claude-mem timeline into per-iso-week files and runs sequential subagents to produce a coherent multi-chapter narrative, with each agent inheriting the prior week's carry-forward block to maintain story continuity.

claude-mem provides persistent agent memory via a live-updating MEMORY.md file and optional real-time observation streaming to messaging channels. setup is automated or manual with clear configuration reference.

knowledge-agent builds filtered observation corpora and loads them into conversational ai sessions for domain-specific querying. users filter observations by project, type, concept, or date, then prime the corpus and ask questions conversationally.

babysit monitors a pull request through its entire lifecycle, polling checks and review threads until all blockers are resolved and the pr is mergeable.

make-plan orchestrates phased implementation planning by delegating fact-gathering to subagents while retaining synthesis. enforces source-backed reporting to ground recommendations in actual apis and code.

how-it-works explains claude-mem's observation capture, memory injection timing (second session onward), and local storage model. covers optional /learn-codebase fast-path and data residency guarantees.

claude-code-plugin-release coordinates semantic versioning and manifest updates across multiple plugin formats and package managers, requiring manual analysis and exhaustive file synchronization.

wowerpoint converts documents into slide deck pdfs via notebooklm, with narrow scope covering only slide generation and explicit exclusion of video and podcast outputs.

learn-codebase reads every source file in a project sequentially to build deep understanding. intended as initial project onboarding or when explicitly requested. lacks specifics on file discovery, ordering, and success criteria.

design-is orchestrates dieter rams audits by delegating evidence collection to subagents, scoring ten principles, and handing off verdicts to /make-plan. comprehensive framework but abstract and untested against real designs.

Map a codebase into feature-grouped flowcharts, identify duplicated concerns across features, and propose a unified architecture. Use when asked to "find the…

Generate a "Journey Into [Project]" narrative report analyzing a project's entire development history from claude-mem's timeline. Use when asked for a timeline…

Execute a phased implementation plan using subagents. Use when asked to execute, run, or carry out a plan — especially one created by make-plan.

Token-optimized structural code search using tree-sitter AST parsing. Use instead of reading full files when you need to understand code structure, find…

Cluster a GitHub issue backlog by root cause into a small set of plan-master issues, redirect children with a standardized comment, and bundle…