Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/petejm/1password-skill/1passwordgit clone --depth 1 https://github.com/petejm/1password-skillWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.08182 | $0.08182 |
| Opus 5 | $0.04091 | $0.04091 |
| Sonnet 5 | $0.01636 | $0.01636 |
| Haiku 4.5 | $0.00818 | $0.00818 |
Grade B, and why
1password scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
Claude Code's bubblewrap sandbox strips setgid bits; `op` requires the `onepassword-cli` group. Symptom: works in terminal but fails as a Claude tool call. Fix: `sudo usermod -aG onepassword-cli $(whoami)`, then log out Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
**Basic auth** (username+password, not token) — never use `curl -u` (mangles special chars): How it starts
The opening of the file, as written. The whole thing — 569 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Requires:
opCLI 2.30.0+ with desktop app integration enabled — 2.30.0 made human-readable output concealed by default and added the--revealflag this skill depends on. Runop --versionto check.
1Password CLI — Decision Router
| You're seeing... | Go to |
|---|---|
| "Permission denied (publickey)" / git push auth fails | → Auth Recovery |
| Need a secret value at runtime | → Runtime Access |
| Don't know the vault, item, or field name | → Runtime Access |
| Setting up op:// in config files / .env | → Config File Patterns |
| SSH agent not responding / wrong socket path | → SSH Agent |
| Git commit signing setup or failures | → Git Signing |
gh, aws, or claude should authenticate through 1Password |
→ Shell Plugins |
| Headless / CI / no desktop app available | → Service Accounts |
op hangs, times out, or biometric won't appear |
→ Troubleshooting |
| Wrong account / multi-account confusion | → Troubleshooting |
op not found / not installed |
→ Troubleshooting |
| Specific error message | → Error Catalog |
Security Rules
- Prefer
op runoverop read— secret never enters conversation context. Never add--no-masking. - Never run
opcommands speculatively — a command that needs authentication raises a biometric prompt on the user's desktop and physically interrupts them. An already-unlocked session answers silently (op signin --help: "'op signin' is idempotent. It only prompts for authentication if you aren't already authenticated"), so you cannot tell in advance which one you will get. Runoponly when the user's request requires it. - Scope with
--vaultonop item get/op item list— prevents exposing item names across all vaults.op readandop runhave no--vaultflag; for those the vault is the first segment of theop://reference. - Never store secrets in files — use
op run --env-fileorop injectfor templating. - Don't bypass security hooks — if a hook blocks
.1password/paths, fix the approach (setSSH_AUTH_SOCKin shell profile, not per-command). - 1Password's Secure AI access guidance: strip hardcoded secrets out of AI/MCP tool configs and let
op runsupply them at launch. See the MCP server pattern under Config File Patterns.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 569 lines · 8,182 tokens per session scan B ca2cc3c74d4d
1password is a cursor rule published in the GitHub repository petejm/1password-skill (4 stars, last pushed 24d ago), licensed Apache-2.0. It adds 8,182 tokens to every session, about $0.0409 per session on Opus 5. A static security scan graded it B with 2 findings (asks for root, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.
control-plane-descriptors
Control plane descriptor and instance implementation patterns.
family-instance-domain-actions
Family instance domain action implementation patterns.