agentleFS
wpultimatesecuritywpultimatesecurity/WordPress-Security-Skills/skills/secrets-credentials-management/SKILL.md

secrets-credentials-management

Skill45 starsChanged 54 days agoMITView on GitHub

Use when handling passwords, API keys, tokens, or third-party credentials in a WordPress plugin or theme. Hashes passwords with wp_hash_password / wp_check_password, generates tokens with wp_generate_password, keeps secrets out of code and the database in plaintext, and uses Application Passwords for API auth. Prevents credential leakage and insecure storage.

  • Reads credentials

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill whenever code handles sensitive credentials:

    line 19
  2. Very specific:

    Revoke what rotation cannot cover: Delete leaked Application Passwords (Users → Profile → Application Passwords) and destroy sessions for …

    line 67
  3. Distinctive:

    Consider history cleanup after revocation, never instead: If needed, plan a git filter-repo rewrite with repository owners and …

    line 76

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Secrets & credentials management

When to use this skill

Use this skill whenever code handles sensitive credentials:

  • Storing an API key, webhook secret, or service token from a settings form.
  • Building a custom login or token-verification flow.
  • Generating reset tokens, API keys, or one-time nonces that must stay secret.
  • Logging or exporting data that might accidentally include secrets.
  • Deciding whether to use the user's account password or an application-specific password for API/integration access.

Secrets in source code, reversible password storage, and plaintext API keys in options are common sources of credential leaks.

Related: see the nonces-csrf-protection skill for CSRF tokens and the settings-options-security skill for sanitizing option values.

Core principles (and why they matter)

  1. Never hardcode secrets. API keys and passwords committed to source control leak to anyone with repo access and often end up in public repositories.
  2. Never store user passwords reversibly. WordPress stores passwords as salted hashes via wp_hash_password(); use wp_check_password() to verify. Do not roll your own hash.
  3. Prefer wp-config constants or encrypted option storage for service keys. Put keys in wp-config.php constants or encrypt them at rest with libsodium so the database alone is not enough to recover them.
  4. Use Application Passwords for API auth. WordPress 5.6+ provides WP_Application_Passwords for machine-to-machine auth without exposing the user's account password.
  5. Never echo or log secrets. Tokens and keys should never appear in HTML, logs, error messages, shell history, or exported backups.
  6. Generate tokens with wp_generate_password() or wp_create_nonce(). They are designed for cryptographic strength (nonces are short-lived; passwords/tokens can be longer).

Step-by-step implementation

  1. For user passwords: hash with wp_hash_password() on save, verify with wp_check_password().
  2. For service/API keys:
    • Accept via a secure settings form (nonce + capability + sanitize).
    • Store in a wp-config.php constant if possible, or encrypt with sodium_crypto_secretbox() using a key derived from wp_salt() (or a dedicated constant) before update_option().
  3. For machine auth: use WP_Application_Passwords::create_new_application_password().
  4. For tokens: generate with wp_generate_password( $length, false ); store only a hash if you need to verify it later.
  5. Sanitize any log/export output to redact known secret keys.
  6. If a secret leaks, respond in this order:
    1. Rotate first. Assume compromise the moment the secret left your control — once pushed to a remote, treat it as captured (clones, forks, and scrapers may already hold it). Issue a new key at the provider, then update the wp-config constant or re-save the encrypted option, and revoke the old key; issuing a replacement alone may leave the leaked key usable. For active abuse, revoke immediately.
    2. Revoke what rotation cannot cover. Delete leaked Application Passwords (Users → Profile → Application Passwords) and destroy sessions for affected users with WP_Session_Tokens::get_instance( $affected_user_id )->destroy_all() for each trusted, verified affected user ID. wp_destroy_all_sessions() targets only the current user, not an arbitrary affected user. Session revocation does not revoke Application Passwords. If wp-config.php leaked, rotate all eight keys/salts and exposed database/service credentials. If an encryption key changes, re-encrypt retained secrets before discarding it; salt-derived keys also change when salts rotate. See wp-hardening-best-practices.
    3. Consider history cleanup after revocation, never instead. If needed, plan a git filter-repo rewrite with repository owners and collaborators. Do not automatically rewrite history or force-push: publishing rewritten history is destructive and requires explicit approval and coordinated protection of others' work. Old clones still contain the secret and can reintroduce it; arrange re-cloning or careful cleanup. Forks and cached views need separate coordination; consult GitHub Support's removal criteria where applicable.
    4. Add push protection to block supported secrets when pushing to GitHub, not when making local commits. Use local secret scanning/pre-commit checks for earlier feedback; neither catches every secret or replaces safe handling.

Supporting references

ReferenceLoad when
Secrets & credentials management checklistBefore final verification of the secrets & credentials management controls.
Secure secret storageImplementing service API-key storage through wp-config.php constants or encrypted options.

Common AI mistakes / anti-patterns

Mistake 1 — API key as a string literal

// ❌ Insecure: key is in source control and readable by anyone with file access.
$api_key = 'sk-live-abc123';
// ✅ Secure: read from a wp-config constant or an encrypted option.
if ( ! defined( 'MY_PLUGIN_API_KEY' ) ) {
    wp_die( esc_html__( 'API key is not configured.', 'my-plugin' ) );
}
$api_key = MY_PLUGIN_API_KEY;

Mistake 2 — Storing user passwords in plaintext or with a weak hash

// ❌ Insecure: reversible or weak storage.
update_option( 'my_plugin_user_password', $_POST['password'] );
update_option( 'my_plugin_user_password', md5( $_POST['password'] ) );
// ✅ Secure: use WordPress password hashing.
$hash = wp_hash_password( $_POST['password'] );
update_option( 'my_plugin_user_password_hash', $hash );

// Verify later:
if ( wp_check_password( $submitted, $stored_hash ) ) { /* ... */ }

Mistake 3 — Logging tokens or API keys

// ❌ Insecure: secret lands in error_log / shell history.
error_log( 'Request failed with token: ' . $api_key );
// ✅ Secure: log an identifier or redacted value.
error_log( 'Request failed for token ending in: ' . substr( $api_key, -4 ) );

Mistake 4 — Reusing nonces as long-lived secrets

// ❌ Insecure: nonces expire and are not designed for persistent auth.
$token = wp_create_nonce( 'my_api' );
update_user_meta( $user_id, 'my_api_token', $token );
// ✅ Secure: generate a dedicated token and hash it for storage.
$token      = wp_generate_password( 32, false );
$token_hash = wp_hash_password( $token );
update_user_meta( $user_id, 'my_api_token_hash', $token_hash );
// Hand the plaintext token to the user once, then verify with wp_check_password().

Mistake 5 — Storing service keys in plaintext options

// ❌ Insecure: database dump exposes the key.
update_option( 'my_plugin_stripe_key', $_POST['stripe_key'] );
// ✅ Secure: encrypt at rest with libsodium (PHP 7.2+).
function my_plugin_encrypt_secret( $plaintext ) {
    $key    = sodium_base642bin( MY_PLUGIN_ENCRYPTION_KEY, SODIUM_BASE64_VARIANT_ORIGINAL );
    $nonce  = random_bytes( SODIUM_CRYPTO_SECRETBOX_NONCEBYTES );
    $cipher = sodium_crypto_secretbox( $plaintext, $nonce, $key );
    return base64_encode( $nonce . $cipher );
}

function my_plugin_decrypt_secret( $encoded ) {
    $raw   = base64_decode( $encoded );
    $nonce = substr( $raw, 0, SODIUM_CRYPTO_SECRETBOX_NONCEBYTES );
    $cipher = substr( $raw, SODIUM_CRYPTO_SECRETBOX_NONCEBYTES );
    $key    = sodium_base642bin( MY_PLUGIN_ENCRYPTION_KEY, SODIUM_BASE64_VARIANT_ORIGINAL );
    return sodium_crypto_secretbox_open( $cipher, $nonce, $key );
}

Correct code examples

A complete settings-page snippet that stores an API key either as a wp-config constant or encrypted in options is in references/secure-secret-storage.php.

Checklist

  • No secrets, API keys, or passwords are hard-coded in source files.
  • User passwords are hashed with wp_hash_password() and verified with wp_check_password().
  • Service/API keys are stored in wp-config.php constants or encrypted at rest.
  • Machine-to-machine auth uses Application Passwords where possible.
  • Tokens are generated with wp_generate_password() (or wp_create_nonce() for short-lived CSRF).
  • Stored tokens are verified against a hash, not compared in plaintext.
  • Secrets are never echoed, logged, exported, or sent to the browser.
  • Leaked credentials are revoked/rotated; affected users' sessions are revoked when warranted.
  • Any history rewrite/publication has explicit approval and coordinated clone/fork cleanup.
  • Settings forms that collect secrets use nonce + capability + HTTPS.

Official references

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.