# 1password > Plan, validate, and use 1Password CLI setup for secret injection and auth. Use when tasks need 1Password CLI usage, secret references, op run/read/inject, or provisioning secrets via env vars/.env files and scripts. Use when this capability is needed. - Skill: `tomevault-io/1password-11` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/1password-11` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/1password-11/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/1password-11 --- # 1Password CLI Use 1Password CLI safely for sign-in, secret references, env injection, and script-driven secret access without exposing secret material. ## Standards snapshot (March 2026) - Follow current 1Password CLI docs for exact command behavior instead of guessing flags or install steps. - Prefer secret references, `op run`, and `op inject` over copying secrets into files or shell history. - Treat app sign-in state, shell session state, and tmux session state as explicit prerequisites. - Keep automation and human flows separate: use service accounts or Environments for unattended use, not ad hoc interactive login hacks. ## When to use - Setting up or validating 1Password CLI. - Reading secrets through secret references. - Injecting secrets into env files, templates, or scripts. - Running commands with secrets via `op run`. ## When not to use - Managing secrets by copying plaintext values into source-controlled files. - Doing generic credential management with no 1Password CLI or secret-reference workflow. - Running `op` directly in unstable shell contexts when tmux-backed sign-in is required. ## Required inputs - The secret-loading goal: sign-in, read, run, inject, or environment provisioning. - OS and shell context. - Account or vault context when multiple accounts are present. - Any target files, templates, scripts, or env workflows involved. ## Deliverables - A safe 1Password CLI path for the requested flow. - Exact prerequisites and verification steps. - Relevant reference-file pointers for the chosen command family. - If requested, a structured status report with a `schema_version` field. ## Philosophy - Secrets should flow through references and runtime injection, not through copied plaintext. - The safest working path is better than the shortest-looking command. - Verification matters: if sign-in or secret resolution is not proven, the workflow is not ready. ## Constraints - Redact secrets, tokens, secret references containing sensitive identifiers, and account details by default. - Do not paste secret values into chat, logs, shell history, code, or committed files. - Do not run `op` outside the required tmux session pattern when interactive sign-in stability matters. - Prefer 1Password Environments or service accounts for automation rather than reusing human interactive auth. ## Workflow 1. Verify OS, shell, and CLI availability with `op --version`. 2. Confirm the desktop app integration or service-account path that fits the request. 3. Start a fresh tmux session for interactive `op` work when sign-in persistence matters. 4. Verify `op signin` and `op whoami` before attempting any secret read. 5. Choose the right command family: - `op run` for env-backed command execution - `op read` for a single secret - `op inject` for templates or config files - `op plugin run` for shell-plugin-assisted flows 6. Return the exact next steps and the reference file for that lane. ## tmux requirement When interactive sign-in is required, use a fresh tmux-backed session and keep all `op` commands inside it. This avoids re-prompts and broken auth state across short-lived shells. ## Tooling and references - Core references: - `references/get-started.md` - `references/cli-examples.md` - `references/cli-reference.md` - `references/secret-references.md` - `references/secrets-environment-variables.md` - `references/secrets-scripts.md` - `references/environment-variables.md` - `references/best-practices.md` - `references/commands-run.md` - `references/commands-read.md` - `references/commands-inject.md` - `references/commands-signin.md` - `references/commands-whoami.md` - `references/contract.yaml` - `references/evals.yaml` - Use assets only when the task needs bundled 1Password support material from `assets/`. ## Validation - Verify CLI availability. - Verify sign-in state with `op whoami` before secret access. - Verify the selected secret-loading pattern matches the requested outcome. - Fail fast at the first missing auth, vault, or session prerequisite. ## Anti-patterns - Running `op` commands in arbitrary short-lived shells after interactive sign-in. - Writing resolved secrets to disk when `op run` or `op inject` would avoid it. - Mixing human interactive auth with unattended automation requirements. - Treating missing sign-in or account context as a minor warning. ## Examples - Help me use `op run` with this `.env` template. - Verify my 1Password CLI setup and tell me why `op whoami` fails. - Show the safest way to inject secret references into this config file. ## See Also | Skill | When to use together | |---|---| | [[security-best-practices]] | Apply secret management best practices alongside 1Password | | [[create-auth]] | Inject credentials into authentication flows via 1Password | | [[cloudflare-deploy]] | Inject secrets into Cloudflare deployments via op run | | [[workers-mcp]] | Secure MCP server secrets with 1Password injection | | [[bootstrap]] | Provision dev environment secrets via 1Password during bootstrap | **Topic map:** [[security-ops]] ## Remember If the workflow would expose a secret unnecessarily, it is the wrong workflow. ## Gotchas - None yet. Capture recurring failures here as symptom -> cause -> do instead -> check. ## Failure mode - If 1Password auth, vault access, or secret references cannot be verified, stop, report the failing check, and fall back to auth/session repair before attempting secret operations. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/jscraik) — claim your Tome and manage your conversions.