inbox-memory

Name: inbox-memory
Author: microsoft

Let AI agents read and write memory with environment-aware storage fallback.

Stars

★ 24

Source

GitHub

Updated

2026-07-08

// securityLow risk

Prompt only, no code
Open source

Running security audit…

Credentials
Network egress
Code execution
Data access
Supply chain

// install

Copy the install command and let the AI configure it · recommended for beginners

Please install the "inbox-memory" skill from askskill:
1. Download https://raw.githubusercontent.com/microsoft/vscode-team-kit/main/github-inbox/skills/inbox-memory/SKILL.md
2. Save it as ~/.claude/skills/inbox-memory/SKILL.md
3. Reload skills and tell me it's ready

// download

Download SKILL.mdInstall manifest ↗

// examples

Store long-term agent preferences

Prompt

Write the following into agent memory: 'Reply in Chinese by default; give the conclusion before steps; clearly label assumptions when information is uncertain.' Then read it back and confirm it was saved.

Expected

A success message plus the saved memory content or a read-back confirmation.

Maintain system rules

Prompt

Add a rule for this agent: 'Mask customer data before processing.' If a similar rule already exists, update it to the latest version and list the current rule set.

Expected

A result showing the rule was added or updated, followed by the cleaned rule list.

Read memory across environments

Prompt

Read the agent's most recently saved project context. If the current environment cannot access the primary storage, automatically fall back to an available backup and state which storage source was used.

Expected

The project context content, plus an indication of whether it came from primary or fallback storage.

// docs

Inbox Memory

This skill reads and writes persistent memory and rules files for the Inbox agent. It abstracts storage so all agents use a single skill instead of calling #memory directly.

File names

Memory: github-inbox-memory.md — session context, patterns, preferences
Rules: github-inbox-rules.md — user's rules and preferences

Storage strategy

Try the #memory tool first (available in VS Code Chat). If #memory is not available (e.g. Copilot CLI), fall back to the file system under ~/.copilot/:

~/.copilot/github-inbox-memory.md
~/.copilot/github-inbox-rules.md

How to detect which backend to use

Attempt #memory with { "command": "view", "path": "/memories/github-inbox-memory.md" }.
If it succeeds or returns "file not found" — #memory is available. Use it for all reads and writes.
If the tool itself is unavailable (tool not found / not supported error) — switch to the file-system fallback for the rest of the session.

Reading

Via #memory:

{ "command": "view", "path": "/memories/github-inbox-memory.md" }
{ "command": "view", "path": "/memories/github-inbox-rules.md" }

Via file system (fallback): Read ~/.copilot/github-inbox-memory.md or ~/.copilot/github-inbox-rules.md using #read (VS Code) or #view (CLI). If the file doesn't exist, treat as empty / first-time user.

Writing

Via #memory:

{ "command": "delete", "path": "/memories/github-inbox-memory.md" }
{ "command": "create", "path": "/memories/github-inbox-memory.md", "file_text": "<content>" }

If the file doesn't exist yet, skip the delete and just use create.

Via file system (fallback):

Ensure ~/.copilot/ exists: mkdir -p ~/.copilot
Write the file using #edit or via terminal: write full content to ~/.copilot/github-inbox-memory.md.

Rules

Always read before writing to avoid losing existing content
Memory and rules are separate files — don't mix them
Keep memory under 100 lines — deduplicate, keep last 5 session logs, be concise
NEVER use repo memory (/memories/repo/) — only the files listed above