Work with macOS Keychain generic password items through bundled scripts. Use when Codex needs to check, read metadata, create, update, retrieve, reveal, or delete local credentials by `service + account` on macOS, especially as a thin credential-store wrapper inside another workflow.
Scanned 9/12/2026
Install to Claude Code
npx -y skills add aibot88/sec_skill_store --skill mac-keychain --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Mac Keychain?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/aibot88-mac-keychain)More formats (shields.io, HTML) on the badges page.
---
name: mac-keychain
description: Work with macOS Keychain generic password items through bundled scripts. Use when Codex needs to check, read metadata, create, update, retrieve, reveal, or delete local credentials by `service + account` on macOS, especially as a thin credential-store wrapper inside another workflow.
metadata:
name: Mac Keychain
description: Work with macOS Keychain generic password items for local credential checks, updates, retrieval, reveal, and deletion.
author: Flc゛
created: 2026-03-27T07:51:16Z
---
# Mac Keychain
Use this skill to operate on macOS Keychain generic password items.
## Scope
- Support `exists`, `read-meta`, `set`, `update`, `read-secret`, `reveal-secret`, and `delete`.
- Operate on generic password items only.
- Stop if the request involves certificates, identities, private keys, or broad keychain administration.
## Runtime
- Require `macOS`, `/usr/bin/security`, and `/bin/zsh`.
- Use the bundled scripts for deterministic behavior.
- Stop and explain the missing runtime dependency if the scripts cannot run as shipped.
## Operating Rules
- Resolve the target with `service + account` before any write, delete, or secret read.
- Default to the current user's `login` keychain.
- Pass `--keychain` only when the target is not the default login keychain.
- Keep caller attribution, approval prompts, and workflow policy in the caller. Do not model them inside this skill.
- Keep the script interface narrow. Do not pass legacy caller-attribution or purpose flags.
- Keep this skill thin. It should expose local Keychain operations, not orchestrate a larger secret-management workflow.
## Safety Model
- Do not guess missing `service` or `account` for `set`, `update`, `read-secret`, `reveal-secret`, or `delete`.
- Do not echo the incoming secret in `set` or `update`.
- Use `read-secret` for downstream machine use after the caller has already decided that retrieval is allowed.
- Use `reveal-secret` only when the user explicitly wants plaintext in visible output. Require `--confirmed`.
- Use `delete` only with a clear target. Require `--confirmed`.
## Scripts
All scripts accept `--service` and `--account`. Use `--keychain` only for a non-default keychain.
- `scripts/keychain-exists.sh`: return whether an item exists.
- `scripts/keychain-read-meta.sh`: return safe metadata only.
- `scripts/keychain-set.sh`: create a new item and fail if it already exists.
- `scripts/keychain-update.sh`: update an existing item and fail if it does not exist.
- `scripts/keychain-read-secret.sh`: return a plaintext secret for downstream use. Use `--value-only` to emit only the secret.
- `scripts/keychain-reveal-secret.sh`: return a plaintext secret for visible output. Require `--confirmed`.
- `scripts/keychain-delete.sh`: delete an item. Require `--confirmed`.
## Output
- Expect structured JSON on stdout for success and on stderr for failures, except `scripts/keychain-read-secret.sh --value-only`.
- Treat `read-secret` output as sensitive even when the caller does not print it.
- Keep normal user-facing status non-disclosing unless the requested action is `reveal-secret`.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!