Memory Bank MCP can read, create, and update files inside the configured memory-bank root., Do not point `MEMORY_BANK_ROOT` at a home directory, source repository root, cloud-sync folder, or any directory containing secrets., Avoid auto-approving write and update tools until the expected file structure and naming rules are established., The path validator rejects path traversal and slash-containing file/project names, but users should still keep the root directory narrow and dedicated., Docker mode requires a bind mount for the memory bank; review the host path and container path before granting write access.
Privacy notes
Memory-bank files can contain project plans, architecture notes, decisions, credentials accidentally pasted into notes, customer context, meeting details, or personal information., Tool calls may expose project names, file names, memory contents, root-path assumptions, and update text to the MCP client and model provider., Multi-project listing can reveal all projects stored under the memory-bank root., Auto-approved memory tools can persist model-generated statements as project memory without human review., Back up important memory-bank directories and review retention expectations before sharing them across clients or machines.
Author
Aliosh Pimenta
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-06
Decision playbook
Review trust signals before you adopt
Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.
Compare context
Selected
0
Current score
63
Baseline
—
Delta
No baseline selected
No major trust-signal divergence detected in the current selection.
Source and provenance checks
Needs review
Confirm ownership and provenance before trusting install instructions.
Source link availableRequired
Open the canonical repository and verify ownership.
Done
Source provenance statusRequired
Marked as source-backed.
Done
Metadata reviewed
No reviewed flag detected in metadata.
Pending
Safety and privacy checks
Complete
Validate risk disclosures before installation or API wiring.
Safety notes presentRequired
Review the listed safety guidance before running commands.
Done
Privacy notes presentRequired
Review data handling notes before connecting accounts or secrets.
Done
Trust level risk gateRequired
Trust level does not block evaluation.
Done
Package and install checks
Needs review
Check package metadata and artifact integrity signals.
Install payload available
Install or copy payload is available for review.
Done
Package verification flag
No package verification flag provided.
Pending
Checksum metadata
No checksum provided for downloaded artifact.
Pending
Compare-driven decision checks
Needs review
Use compare context to validate trade-offs before adoption.
Compare tray has multiple entries
Add at least one more entry to compare trust differences.
5 safety and 5 privacy notes across 5 risk areas. Review closely: credentials & tokens, permissions & scopes, third-party handling.
5 areas
SafetyLocal filesMemory Bank MCP can read, create, and update files inside the configured memory-bank root.
SafetyCredentials & tokensDo not point `MEMORY_BANK_ROOT` at a home directory, source repository root, cloud-sync folder, or any directory containing secrets.
SafetyLocal filesAvoid auto-approving write and update tools until the expected file structure and naming rules are established.
SafetyLocal filesThe path validator rejects path traversal and slash-containing file/project names, but users should still keep the root directory narrow and dedicated.
SafetyPermissions & scopesDocker mode requires a bind mount for the memory bank; review the host path and container path before granting write access.
PrivacyCredentials & tokensMemory-bank files can contain project plans, architecture notes, decisions, credentials accidentally pasted into notes, customer context, meeting details, or personal information.
PrivacyThird-party handlingTool calls may expose project names, file names, memory contents, root-path assumptions, and update text to the MCP client and model provider.
PrivacyData retentionMulti-project listing can reveal all projects stored under the memory-bank root.
PrivacyData retentionAuto-approved memory tools can persist model-generated statements as project memory without human review.
PrivacyData retentionBack up important memory-bank directories and review retention expectations before sharing them across clients or machines.
Disclosure: MIT-licensed TypeScript MCP server inspired by Cline Memory Bank. It manages project memory files from a configured root and is separate from broader knowledge-graph, note-taking, or long-term memory servers in the catalog.
Safety notes
Memory Bank MCP can read, create, and update files inside the configured memory-bank root.
Do not point `MEMORY_BANK_ROOT` at a home directory, source repository root, cloud-sync folder, or any directory containing secrets.
Avoid auto-approving write and update tools until the expected file structure and naming rules are established.
The path validator rejects path traversal and slash-containing file/project names, but users should still keep the root directory narrow and dedicated.
Docker mode requires a bind mount for the memory bank; review the host path and container path before granting write access.
Privacy notes
Memory-bank files can contain project plans, architecture notes, decisions, credentials accidentally pasted into notes, customer context, meeting details, or personal information.
Tool calls may expose project names, file names, memory contents, root-path assumptions, and update text to the MCP client and model provider.
Multi-project listing can reveal all projects stored under the memory-bank root.
Auto-approved memory tools can persist model-generated statements as project memory without human review.
Back up important memory-bank directories and review retention expectations before sharing them across clients or machines.
Prerequisites
Node.js and npm available for `npx`.
A dedicated memory-bank root directory, separate from source repositories and secrets.
Project-specific subdirectories created under `MEMORY_BANK_ROOT`.
MCP client approval rules reviewed before allowing write or update operations.
Optional Docker volume path reviewed if running the server in a container.
Schema details
Install type
cli
Troubleshooting
No
Source repository stats
Scope
Source repo
Collection metadata
Estimated setup
10 minutes
Difficulty
beginner
Tool listing metadata
Disclosure
MIT-licensed TypeScript MCP server inspired by Cline Memory Bank. It manages project memory files from a configured root and is separate from broader knowledge-graph, note-taking, or long-term memory servers in the catalog.
Memory Bank MCP Server turns Cline-style project memory banks into an MCP
service. It stores project-specific memory files under a configured root and
exposes tools for listing projects, listing project files, reading files,
writing new files, and updating existing memory files.
Use it when a team wants a consistent file-backed memory-bank structure across
Claude, Cline, Roo Code, Cursor, or another MCP client. Keep it scoped to a
dedicated memory directory so the agent manages notes rather than arbitrary
local files.
These sources were reviewed on 2026-06-06. Prefer the live repository,
README, npm metadata, license, package metadata, custom instructions, path
validator, MCP route definitions, and filesystem repository implementation for
current setup and behavior details.
Features
Manage memory-bank files through a TypeScript MCP server.
Use MEMORY_BANK_ROOT to define the storage root.
List available projects under the memory-bank root.
List files for one project.
Read memory-bank file contents.
Write new memory-bank files.
Update existing memory-bank files.
Validate required parameters and reject path traversal or slash-containing
file/project names.
Run through npx, Smithery, or Docker with a mounted memory-bank directory.
Installation
Create or choose a dedicated memory-bank directory, then configure the server:
If running with Docker, bind-mount only the memory-bank directory that should be
visible to the server.
Use Cases
Maintain project context files shared by Claude, Cline, Roo Code, and Cursor.
Keep architecture notes, decisions, active tasks, progress summaries, and
project conventions in a consistent memory-bank structure.
Let an agent read memory before starting a task and update memory after a
reviewed milestone.
Manage separate memory directories for multiple projects from one MCP server.
Pair memory-bank updates with human review before writing durable project
state.
Safety and Privacy
Memory Bank MCP is intentionally file-backed. That makes it simple to inspect
and back up, but it also means write tools create durable notes. Use a dedicated
root, do not store secrets in memory files, and keep write/update operations out
of auto-approval until the workflow is trusted.
Review memory files periodically. Project memories can drift, accumulate stale
decisions, or accidentally preserve sensitive context that should have stayed in
the chat session.
Duplicate Notes
Existing catalog entries cover other memory systems, knowledge graphs,
note-taking integrations, and agent memory tools. This entry covers the separate
alioshr/memory-bank-mcp package for file-backed, Cline-style project memory
bank management.
Show that Memory Bank MCP Server is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.
[](https://heyclau.de/entry/mcp/memory-bank-mcp-server)
How it compares
Memory Bank MCP Server side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
MCP server for Trello boards, cards, lists, comments, checklists, attachments, labels, board activity, workspace selection, and rate-limited Trello API automation from Claude.
MCP server and CLI for Ruflo's multi-agent Claude Code and Codex workflow harness, exposing tools for swarms, agents, memory, hooks, daemons, plugins, federation, and project automation.
MCP server for Todoist task and project management, with tools for tasks, subtasks, projects, sections, labels, comments, reminders, filters, activity, backups, collaboration, bulk operations, and dry-run testing.
✓Memory Bank MCP can read, create, and update files inside the configured memory-bank root.
Do not point `MEMORY_BANK_ROOT` at a home directory, source repository root, cloud-sync folder, or any directory containing secrets.
Avoid auto-approving write and update tools until the expected file structure and naming rules are established.
The path validator rejects path traversal and slash-containing file/project names, but users should still keep the root directory narrow and dedicated.
Docker mode requires a bind mount for the memory bank; review the host path and container path before granting write access.
✓Trello MCP Server can create, read, update, move, archive, and delete board data depending on token permissions and requested tools.
Card, checklist, label, due-date, start-date, comment, attachment, board, list, workspace, and active-board changes can affect real team workflows.
File attachments are added from URLs; review source URLs and destination cards before allowing an agent to attach files.
Use `TRELLO_ALLOWED_WORKSPACES` to limit workspace access, especially when a token can see personal, client, or production boards.
Require confirmation before creating boards, adding attachments, deleting or archiving cards, changing due dates, moving cards between lists, or posting comments on behalf of a user.
✓Ruflo's full CLI install can add project files, MCP configuration, hooks, helper scripts, memory state, commands, skills, and background automation.
The MCP surface includes tools for agent spawning, swarm coordination, memory operations, hook execution, daemon control, plugins, security scans, and federation.
Background daemons and hooks can run automatically based on session, task, file, or worker events; inspect generated settings before relying on them.
Some workflows can execute shell commands, edit project files, scan dependencies, start services, or dispatch multi-step agent tasks.
Federation and HTTP transport modes can expose agent coordination outside the local stdio boundary; keep them disabled unless you have reviewed network trust.
Use least-privilege repositories and require human review before accepting generated code, security fixes, plugin installs, or automated workflow changes.
✓Todoist MCP Server can create, update, delete, complete, reopen, reorder, archive, and bulk-modify tasks, subtasks, projects, sections, labels, comments, reminders, filters, shared labels, and collaboration state.
Bulk create, bulk update, bulk delete, bulk complete, project deletion, section deletion, label deletion, reminder changes, and duplicate merge operations can alter many Todoist records at once.
Dry-run mode simulates mutation operations but still queries real Todoist data for validation.
Pro and Business features such as reminders, filters, shared labels, collaboration, backups, and workspace operations may expose or modify team data.
Require confirmation before deleting tasks or projects, changing due dates in bulk, completing groups of tasks, modifying shared projects, creating reminders, downloading backups, or changing collaboration state.
Privacy notes
✓Memory-bank files can contain project plans, architecture notes, decisions, credentials accidentally pasted into notes, customer context, meeting details, or personal information.
Tool calls may expose project names, file names, memory contents, root-path assumptions, and update text to the MCP client and model provider.
Multi-project listing can reveal all projects stored under the memory-bank root.
Auto-approved memory tools can persist model-generated statements as project memory without human review.
Back up important memory-bank directories and review retention expectations before sharing them across clients or machines.
✓Trello API keys, tokens, board IDs, workspace IDs, card IDs, member names, labels, comments, checklists, attachments, due dates, and activity logs can be exposed to the MCP client.
Active board and workspace selections may persist locally in Trello MCP configuration.
Card descriptions, comments, attachments, and activity histories can contain private customer, roadmap, incident, or operational data.
Proxy settings, Trello credentials, and workspace allowlists should be kept out of shared MCP configs, logs, screenshots, and transcripts.
Response previews reduce some large card-description exposure, but full-card and activity tools can still return sensitive Trello content.
✓Ruflo memory, hooks, logs, telemetry-style metrics, worker output, prompts, code snippets, file paths, and task history may be stored under project-local state.
Agent memory and embeddings can preserve sensitive implementation details beyond the current chat session.
Federation, hosted UI, remote provider, plugin, or HTTP transport workflows may move task context beyond the local machine if enabled.
Security scans, dependency checks, and code analysis can reveal private package names, vulnerabilities, paths, branch names, and repository structure.
Review `.claude`, `.claude-flow`, helper scripts, logs, and exported memory before sharing a repository or support bundle.
✓Todoist API tokens, task IDs, project IDs, section IDs, label IDs, reminder IDs, comment IDs, user IDs, collaborator details, activity logs, completed tasks, backups, and productivity statistics can be exposed to the MCP client.
Task names, descriptions, comments, due dates, deadlines, labels, reminders, attachments, project notes, and completed-task history can contain private work or personal planning data.
Bulk and filter tools may reveal larger slices of a Todoist workspace than a single-task request.
Backup download tools can expose a broad historical snapshot of Todoist data.
Keep Todoist tokens in local MCP configuration only, and avoid sharing MCP logs, dry-run traces, or transcripts that include private task or project data.
Prerequisites
Node.js and npm available for `npx`.
A dedicated memory-bank root directory, separate from source repositories and secrets.
Project-specific subdirectories created under `MEMORY_BANK_ROOT`.
MCP client approval rules reviewed before allowing write or update operations.
Trello API key and token with access to the boards and workspaces Claude may use.
Bun-compatible published package runtime or an MCP client that can launch the packaged stdio server through npx.
Optional `TRELLO_BOARD_ID` or `TRELLO_WORKSPACE_ID` when a default board or workspace should be selected at startup.
Optional `TRELLO_ALLOWED_WORKSPACES` to constrain the token to specific approved Trello workspaces.
Node.js 20 or newer.
Claude Code or another MCP client that can launch stdio servers.
A repository where Ruflo-generated `.claude`, `.claude-flow`, settings, helpers, hooks, and memory files are acceptable.
Review of which MCP tool groups, hooks, plugins, providers, and background workers should be enabled.
Todoist API token from Todoist integrations developer settings.
Node.js 18 or newer for the npm stdio server.
Review of whether the token has access to personal, shared, Pro, Business, or workspace features.
Optional `DRYRUN=true` for simulating mutation tools while still validating against real Todoist data.