Add auth token subcommand to output active API token

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.

[jedisct1]

Change summary

This adds a new command: auth token, which prints the currently active token (if there is one).

Unfortunately, it is still common to require a valid token to run curl commands directly, since much of our documentation is centered around the API.

AI agents are especially problematic. If they browse our API documentation, they will often try to obtain a valid Fastly token in order to use curl directly, even though the CLI is installed. In practice, they will use the fastly CLI to retrieve the token and then use it in curl commands.

Even worse, some users may not be running private or local models, which means these tokens could be leaked to random service providers.

Beyond the AI case, people may also want to use these tokens in CI/CD pipelines. Currently, they have to rely on unreliable hacks with grep and sed to extract them.

This PR tries to give both people and AI agents what they need instead of letting them do questionable things.

The fastly auth token command returns the currently active token.

To avoid printing the token directly to a terminal, the command explicitly checks that the output is not a terminal. In other words, it is designed to be used in shell substitution (for example: curl -H "Fastly-API-Token: $(fastly auth token)" ...).

And of course, the token is not accessible if FASTLY_DISABLE_AUTH_COMMAND is set.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

User Impact

Users and agents get a reliable way to use tokens in scripts.

Are there any considerations that need to be addressed for release?

[rcaril]

Some minor suggestions, otherwise looks good.

[rcaril]

Also adding @kpfleming to take a look as well.

[kpfleming]

Looks fine to me. The test files can be combined or not at your discretion.