If you are an LLM, install the CLI, then run agentself install --skills.

agentself gives an agent a persistent local identity: a wallet, encrypted secrets, non-secret notes, and optional email. It is a CLI first, with one command tree across the supported backends.

Linux, macOS, and Windows. Python 3.11+. This is the first stable release. Review COMPATIBILITY.md before automating against the CLI.

uv tool install agentself

Or:

pipx install agentself

If neither tool is available:

python -m pip install --user agentself

agentself install --skills
agentself install --tools
agentself init
agentself wallet address
agentself secret create API_TOKEN --file ./api-token.txt
agentself note set handoff --file ./handoff.txt
agentself email connect

init creates one identity directory. Run agentself show to inspect it; repeat init is safe, and --force is required to change an existing identity or its backends. Email setup is optional and does not block init. For AgentMail, email connect offers two explicit routes: connect an existing API key, or create an account when that external action has been authorized. The account route asks for the approved human email, sends its six-digit OTP, and keeps the generated API key encrypted while setup continues.

Use the CLI for the exact current flags and input rules:

agentself --help
agentself wallet --help
agentself email --help

Secret file input drops a leading UTF-8 BOM and keeps a trailing newline. wallet.key must be a hex private key after that decode. Retrieval is safe by default: use secret get NAME --file PATH or --meta; plaintext stdout requires --print. Human --print ends with one LF; --file is byte-exact.

note set is an idempotent upsert for printable, non-secret handoff context. It accepts VALUE or --file PATH; file input drops a leading UTF-8 BOM and otherwise preserves UTF-8 bytes and newlines. Human note get ends with one LF. Notes are included in identity backup/restore. Never put credentials, OTPs, private keys, secret values, or mail bodies in notes.

email list, email find QUERY, and default email receive print headers and new/seen status without bodies. Surfaced messages include the raw provider id and a stable compact identity-local ref such as m1. Use the ref with email receive REF --file PATH or email mark REF acted|unacted; raw provider IDs remain accepted. acted is independent local task state. The compact syntax is reserved: an unknown matching ref is refused instead of being sent to a provider. Acted state can be filtered with email list --acted|--unacted or the same flags on email find. Bodies can contain API keys and login links; --print is an explicit unsafe choice.

Backends keep the public commands stable while allowing the implementation to change:

agentself backends
agentself backends wallet
agentself backends email
agentself backends store

Choose a backend during initialization:

agentself init --wallet ethereum --email imap --store pass

Wallet and email backends can also be selected with AGENTSELF_WALLET_BACKEND and AGENTSELF_EMAIL_BACKEND. The precedence is CLI flag, environment variable, saved identity configuration, then default; there is no automatic failover. Backend-specific setup, credential sources, and required host settings are exposed by agentself backends and agentself email connect.

The identity directory is ~/.agentself by default. Set AGENTSELF_IDENTITY_DIR to isolate identities, for example:

AGENTSELF_IDENTITY_DIR=~/.agent-a agentself init
AGENTSELF_IDENTITY_DIR=~/.agent-b agentself init

Use AGENTSELF_EMAIL_ADDRESS and AGENTSELF_EMAIL_CREDENTIAL for transient email configuration when appropriate. Runtime environment credentials are not copied into the identity; complete email connect with its secure result file when the credential should be encrypted into the identity. If authorized AgentMail signup reports that the requested identity is claimed, forbidden, or unavailable, stop and use the existing-key route. The CLI does not probe alternate aliases.

The CLI is designed to be discovered from a shell. No MCP server or Python import is required:

agentself --help
agentself backends
agentself diagnose

For automation, prefer --json. Successful and failed commands emit one JSON object; human-readable errors stay on stderr. Exit codes are 0 for success, 1 for an error, 2 for a refusal, and 3 when an input, dependency, or resource is missing. agentself --json --version reports the package version and machine-output schema version. Consumers should ignore unknown JSON keys.

An optional bundled skill provides the same discovery guidance to compatible coding agents:

agentself install --skills
agentself install --skills -g

• Secret values are encrypted at rest by the configured store.
• Notes are non-secret and printable; private file modes are defense in depth, not encryption.
• Secret listings return names, not values; plaintext needs --print, and wallet.key also requires --unsafe.
• Email bodies stay off stdout unless email receive --print is explicit; prefer ID --file PATH.
• Logs and structured errors redact secret values.
• Missing credentials and host tools fail instead of silently switching backends.
• Wallet backends are live; wallet send can move real assets.
• Identity files are replaced atomically, and wallet-send retries do not create a second payment.

Back up an identity with agentself backup PATH. Report vulnerabilities through SECURITY.md.

New backends implement the existing channel contract and do not add provider-specific CLI verbs. See CONTRIBUTING.md.

git clone https://github.com/agentself/agentself
cd agentself
uv sync
uv run pytest
uv run ruff check .
uv run mypy agentself

See RELEASE.md for the small-project release checklist.

MIT