cleanmemory
Reset Max Prompter's project-context memory boundary so unrelated projects and prior-project notes do not influence the current workspace.
Skill by bvts
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/bvts-cleanmemory
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/cleanmemory
What's inside
- Skill name
cleanmemory- Description
- Reset Max Prompter's project-context memory boundary so unrelated projects and prior-project notes do not influence the current workspace. Preserve the current chat and current project files. Use when the user wants to start clean or remove cross-project context.
- License
- MIT
- Instructions
- 101 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# /cleanmemory — CROSS-PROJECT CONTEXT RESET ## Mission Run a safe context-isolation reset for Max Prompter. The purpose is to prevent memories, notes, council transcripts, prompts, or cached context from unrelated projects from influencing work in the current project. This command operates on **Max Prompter's local project-context artifacts and the agent's instructions for the current session**. It does **not** claim to delete model-provider conversation history or hidden platform memory, and it must not attempt to modify the current chat. ## Safety rule Never delete, rewrite, or move ordinary project source files merely because they look like memory. Do not touch: - source code - assets - user documents - databases - configuration needed by the application - `.git` history - the current conversation - current project requirements/specs unless they are explicitly marked as Max Prompter cross-project context ## Procedure ### 1. Identify the current workspace Determine the opened workspace root and treat it as the active project. ### 2. Discover Max Prompter context artifacts Look only for clearly Max-owned or explicitly context-oriented locations, including where present: ```text .max-prompter/ .max-prompter/memory/ .max-prompter/context/ .max-prompter/cache/ .max-prompter/council/ ``` Also inspect obvious Max-generated files whose names explicitly indicate cross-project memory, such as: ```text max-memory* max-context* cross-project-memory* council-transcript-* ``` Do not broadly delete every file named `memory`, `context`, `notes`, or `history` because those names may belong to the project itself. ### 3. Preserve instead of hard-delete When Max-owned context artifacts are found, create a timestamped archive directory such as: ```text .max-prompter/archive/context-reset-YYYYMMDD-HHMMSS/ ``` Move only the clearly Max-owned cross-project context artifacts into that archive. If a safe ownership determination cannot be made, leave the file untouched and report it. ### 4. Establish a clean context boundary Create or update: ```text .max-prompter/CLEAN_CONTEXT.md ``` The file must state that the current project is the sole active project context for this session and that unrelated historical project context must not be used unless the user explicitly reintroduces it. ### 5. Continue from the current project After the reset: - read the current project's own instructions normally; - use the current conversation normally; - use current workspace files normally; - ignore archived cross-project context unless the user explicitly asks to restore it. ## Required user-facing result Report: 1. that the cross-project Max context boundary was reset; 2. what Max-owned context was archived, if any; 3. that the current chat and current project files were preserved; 4. that platform/model conversation history was not deleted by this command. Do not imply that this command can erase hidden model memory or provider-side conversation history. ## Idempotency Running `/cleanmemory` twice without generating new cross-project context should be safe and should produce little or no additional change.
From bvts/max-the-prompter, MIT.
More from bvts/max-the-prompter
- 0.579score
build
bvts
Implements an approved plan in coherent, verifiable slices while preserving existing architecture and behavior.
SkillAFreshness: A★ 1Claude Code +9 - 0.547score
council
bvts
Pressure-tests meaningful decisions through a five-advisor Council using Contrarian, First Principles, Expansionist, Outsider, and Executor perspectives, peer…
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
audit
bvts
Performs a broad project audit across correctness, architecture, UX, accessibility, security, performance, runtime behavior, and maintainability.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
ship
bvts
Prepares and verifies a project for production release, including tests, build, deployment configuration, security, observability, rollback, documentation…
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
design
bvts
Designs or redesigns user interfaces with a coherent visual system, appropriate component sources, motion, responsive behavior, and rendered-browser…
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
review
bvts
Runs a structured pre-merge review of changes for correctness, architecture, readability, maintainability, security, performance, tests, and accidental scope…
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
spec
bvts
Creates a complete, implementation-ready project or feature specification before code.
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
skillinstall
bvts
Install a third-party or local Agent Skill into the current workspace under a clean skills/<skill-name>/ directory, optionally discovering the repository from…
SkillBFreshness: B★ 1Claude Code +9 - 0.542score
prompt
bvts
Generates an execution-ready implementation prompt from a user request or project context using the Max Prompter workflow.
SkillBFreshness: B★ 1Claude Code +9 - 0.542score
constraints
bvts
Establishes and enforces a measurable project quality bar, normally persisted as CONSTRAINTS.md.
SkillBFreshness: B★ 1Claude Code +9 - 0.539score
plan
bvts
Converts a project specification into dependency-ordered, atomic, verifiable implementation tasks.
SkillBFreshness: B★ 1Claude Code +9 - 0.538score
re
bvts
General reverse-engineering and debugging workflow for authorized software.
SkillBFreshness: B★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 23, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/bvts/max-the-prompter
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bvts
- 0.579score
build
bvts
Implements an approved plan in coherent, verifiable slices while preserving existing architecture and behavior.
SkillAFreshness: A★ 1Claude Code +9 - 0.547score
ship
bvts
Prepares and verifies a project for production release, including tests, build, deployment configuration, security, observability, rollback, documentation…
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
audit
bvts
Performs a broad project audit across correctness, architecture, UX, accessibility, security, performance, runtime behavior, and maintainability.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
council
bvts
Pressure-tests meaningful decisions through a five-advisor Council using Contrarian, First Principles, Expansionist, Outsider, and Executor perspectives, peer…
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
design
bvts
Designs or redesigns user interfaces with a coherent visual system, appropriate component sources, motion, responsive behavior, and rendered-browser…
SkillBFreshness: B★ 1Claude Code +9 - 0.544score
review
bvts
Runs a structured pre-merge review of changes for correctness, architecture, readability, maintainability, security, performance, tests, and accidental scope…
SkillBFreshness: B★ 1Claude Code +9
Cite this page
cleanmemory skill. Agent Mart. https://agentmart.dev/skills/bvts-cleanmemory (last updated 2026-10-24)