Your secrets don't need a git commit to leak into an AI tool
Post · https://dev.to/cyclopt_dimitrisk/your-secrets-dont-need-a-git-commit-to-leak-into-an-ai-tool-be2
Most of us have the "don't commit secrets" reflex drilled in. Pre-commit hooks, secret scanners in...
First seen by Oriel on .
AI summary
The article argues that AI coding assistants expand secret exposure beyond repositories because pasted text, automatically gathered context, and agent-accessible files can carry credentials to external services.
- Secrets can enter AI context through failing tests that print configuration, environment dumps, verbose logs, or snippets copied from Slack.
- The author recommends minimizing pasted context, redacting credentials, using short-lived least-privilege keys, checking retention and telemetry settings, and restricting agent access to secret files.
- Deleting a chat message may not remove data already logged, cached, or forwarded; exposed credentials should be rotated and revoked, with access logs checked.
- The examples are illustrative rather than reports of specific incidents.
#ai coding assistants #secrets management #data security
Generated by gpt-5.6-sol. Check the original for details.
Metric trends
Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained
Reactions
Total public reactions reported by the source; distinct from points and likes.
- DEV Community: 74, latest observation ; +15 over observed period since
Comments
Comment count reported by the source.
- DEV Community: 118, latest observation ; +41 over observed period since
Aliases
| Source | External ID | URL |
|---|---|---|
| DEV Community | 4921112 | https://dev.to/cyclopt_dimitrisk/your-secrets-dont-need-a-git-commit-to-leak-into-an-ai-tool-be2 Read discussion |