Skip to main content

Persistent memory for Claude Code — local or VPS

Project description

SecBrain

Persistent memory for Claude Code — local or VPS.

SecBrain gives Claude Code a long-term memory vault that persists across sessions. Install it once globally, run secbrain init in any project, and Claude will automatically remember bugs, features, decisions, and learned patterns.


Install

pip install secbrain -q

Quick Start

# In any project directory
cd my-project
secbrain init

The wizard will:

  1. Ask where your vault lives (local ~/.secbrain/vault or remote VPS)
  2. Deploy the MCP server (locally or via SSH)
  3. Register the current project
  4. Inject CLAUDE.md instructions and .mcp.json MCP config
  5. Install Claude Code session hooks

Then start a Claude Code session — memory is active automatically.


Commands

Command Description
secbrain init Full setup wizard
secbrain connect [name] Connect current directory to a vault project
secbrain projects List all vault projects
secbrain status Show server health and current project
secbrain start Start local vault server
secbrain stop Stop local vault server
secbrain logs [project] Show session log
secbrain bugs [project] Show known bugs
secbrain deploy (Re)deploy server to VPS
secbrain eject Remove SecBrain from current project
secbrain skills List learned skills
secbrain skills search <query> Search skills
secbrain skills show <name> Show a skill

What Gets Stored

~/.secbrain/vault/
├── _system/
│   ├── log.md          ← all session start/end events
│   └── entities.db     ← SQLite knowledge graph
├── projects/{name}/
│   ├── overview.md     ← project summary
│   ├── bugs.md         ← logged bugs
│   ├── features.md     ← logged features
│   ├── decisions.md    ← architectural decisions
│   ├── sessions.md     ← session summaries
│   └── user_model.md   ← learned user preferences
├── skills/
│   ├── global/         ← cross-project reusable patterns
│   └── {project}/      ← project-specific patterns
├── raw/                ← knowledge ingestion staging
└── wiki/               ← processed topic wikis

MCP Tools Available to Claude

Tool When to call
vault_start_session(project, cwd) Automatically via hook at session start
vault_end_session(project, summary) At end of session
vault_log_bug(project, title, description, severity) When discovering a bug
vault_log_feature(project, title, description, status) When implementing a feature
vault_log_decision(project, title, rationale) When making an arch decision
vault_recall(project, query) Search everything
vault_update_overview(project, overview) Update project description
vault_save_skill(...) After solving something non-obvious
vault_recall_skill(query, project) Before starting complex tasks
vault_improve_skill(project, skill_name, refinement) When a skill gets better
vault_update_user_model(project, observation) When noticing user patterns
vault_read(path) Read any vault file
vault_write(path, content) Write any vault file
vault_list(path) List vault directory
vault_list_projects() List all projects

VPS Deploy

secbrain init
# Select "VPS" → enter user@host → confirm deploy

SecBrain installs itself on the VPS via pip and runs as a systemd service on port 8765.


Multiple Projects

cd ~/other-project
secbrain connect
# Select existing project or create new

Development

git clone ...
cd secbrain
pip install -e ".[dev]"

Requirements

  • Python 3.10+
  • Claude Code with MCP support
  • For VPS: SSH access + pip on remote

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

secbrain-0.1.5.tar.gz (28.7 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

secbrain-0.1.5-py3-none-any.whl (33.0 kB view details)

Uploaded Python 3

File details

Details for the file secbrain-0.1.5.tar.gz.

File metadata

  • Download URL: secbrain-0.1.5.tar.gz
  • Upload date:
  • Size: 28.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for secbrain-0.1.5.tar.gz
Algorithm Hash digest
SHA256 cdcc240fc09d66a66e52db9bb4ae1d30a7cbd1e0ecf3bc78c748e16730fc33a2
MD5 1f46092e61ae7e0a194f7c580b5d008b
BLAKE2b-256 4601dd82f4aa5f8dc44028b33a6e3c5ddad501cb214d5d3ef1bbd0a887ac5803

See more details on using hashes here.

Provenance

The following attestation bundles were made for secbrain-0.1.5.tar.gz:

Publisher: publish.yml on KoodosBots/SecBrain

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file secbrain-0.1.5-py3-none-any.whl.

File metadata

  • Download URL: secbrain-0.1.5-py3-none-any.whl
  • Upload date:
  • Size: 33.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for secbrain-0.1.5-py3-none-any.whl
Algorithm Hash digest
SHA256 fbc4ab86dcbb0ac871713593e1eefb8f9f86b79436a5baf3a609e290d4c4e1e4
MD5 7326b6280e2d290d0e8931371ec619c2
BLAKE2b-256 dc066a3319cc510fdb09666cd1f73849812aa8e8f207fd0d5f9baa5ce1289ed7

See more details on using hashes here.

Provenance

The following attestation bundles were made for secbrain-0.1.5-py3-none-any.whl:

Publisher: publish.yml on KoodosBots/SecBrain

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page