Stop telling agents to ask for the user's Sandbox secret - #40
Merged
Merged
Conversation
All three guides opened by instructing the agent to ask the developer for their client_id and secret when not provided. An agent following that puts a live credential into a model's context and into the transcript. The Plaid CLI can fetch them from the Dashboard and write the .env itself, so the guides now point at `plaid keys write --env sandbox` and say plainly not to ask for a secret in chat. The Prerequisites line is updated the same way, since it assumed the credentials had been copied by hand.
phoenixy1
approved these changes
Sep 10, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
All three guides open with the same assumption:
An agent that follows this asks the developer to paste a live Plaid secret into a chat window — so into a model's context, into the transcript, and into whatever retains that transcript. It's the exposure these guides otherwise warn about, and right now it's the documented behavior.
There's a better path than there was when this was written. The Plaid CLI fetches credentials from the Dashboard and writes the
.envitself, so nobody has to handle the secret:That shipped in the CLI release on 2026-09-09, so the guides can point at it directly. Each one now does, and says plainly not to ask for a secret in chat. The Prerequisites line gets the same treatment — it assumed the credentials had already been copied by hand from the Dashboard.
--env sandboxis explicit on purpose: without it the CLI writes whichever environment it's pointed at, which is Production for teams that have one.One line per file plus the prerequisite, and no new files — the same guidance is published as an agent skill at
plaid.com/.well-known/skills/, so this defers to the CLI rather than restating it.