Data handling risk

Customer Data in Agent Context

Support exports, logs, database samples, transcripts, and screenshots can enter agent context during normal debugging work.

Severity: HighDetected by: Growth and Pro

Signals CapitalGuard looks for

Support exports committed near application code

Logs or fixtures with real customer identifiers

Screenshots, transcripts, or CSV files used for debugging

Agents can summarize or transform data into new places faster than humans notice.

Debugging artifacts often survive longer than intended.

Customer data exposure can create contractual, reputational, and operational risk.

Use synthetic fixtures instead of real customer records.

Block customer-data paths from agent reads and uploads.

Delete temporary support exports after the authorized review is complete.

Next step

Related threats

Agent instruction risk

Hidden instructions in docs, issues, prompts, logs, and internal notes can influence coding agents when those files become task context.

AI tools do not need to print raw keys to create risk; file paths, variable names, logs, and configuration context can reveal how systems are wired.

Some files should not be read, edited, summarized, or sent to external tools by autonomous coding agents.