Social·Community
Developers weigh secrets managers for AI agents without exposing credentials
A Reddit thread asks how practitioners handle secrets and sensitive credentials when running agents. The poster notes ad hoc approaches like secrets.yaml files and SSO-based auth, and asks whether purpose-built, agent-friendly secrets management tools exist. The thread reflects a gap between agent adoption and established secrets-handling practice.
30 engagement1 sourceSat, Oct 3, 2026, 04:06 PM
AI Overview
- A Reddit thread asks which password or secrets managers are agent-friendly for handling credentials and sensitive data.
- Respondents cited approaches including secrets.yaml files and SSO-based authentication for agent access to secrets.
- The poster seeks solutions purpose-built for agents rather than ad hoc secrets storage.
Entities
Related
Products·launched
kdbx run: inject KeePassXC secrets into agent shell commands
kdbx run is an open-source CLI tool that injects secrets from a KeePassXC vault into child processes without printing them, so AI agents with shell access can run commands needing credentials without exposing them in transcripts. It solves the problem of API keys leaking when agents read .env files, giving developers control over secret access.
kdbx run is an open-source CLI tool that injects secrets from a KeePassXC vault into child processes without printing them, so AI agents with shell access can run commands needing credentials without exposing them in transcripts. It solves the problem of API keys leaking when agents read .env files, giving developers control over secret access.
35 engagement1 sourceredditSat, Aug 1, 2026, 01:25 AM
Social·Community
How do you handle API keys when an AI agent needs access to multiple external services?
Developer asks how to manage API keys for AI agents calling multiple external services
Developer asks how to manage API keys for AI agents calling multiple external services
9 engagement1 sourceredditThu, Sep 24, 2026, 08:11 AM
Products·launched
Argus scans AI agent sessions for leaked credentials
Argus is a security tool that monitors AI coding agent sessions to detect when secrets like API keys or passwords are exposed. It helps developers identify which sessions used a secret and how, enabling them to respond to potential leaks.
Argus is a security tool that monitors AI coding agent sessions to detect when secrets like API keys or passwords are exposed. It helps developers identify which sessions used a secret and how, enabling them to respond to potential leaks.
1 engagement1 sourceredditWed, Aug 19, 2026, 08:27 PM
Releases·Tool Release
Show HN: MagicVault – Make credentials usable by agents yet hidden from them
Developer launches MagicVault to hide credentials from AI agents while keeping them usable
Developer launches MagicVault to hide credentials from AI agents while keeping them usable
4 engagement2 sourceshackernewsWed, Sep 9, 2026, 02:08 PM
Social·Community
What should a coding agent be allowed to do?
A Reddit discussion asks how teams constrain what coding agents can access and do inside projects. Agents often inherit broad permissions, including `.env` files with API keys and database credentials, and can install dependencies, call APIs, open PRs, or deploy. The poster asks whether teams use short-lived credentials, scoped service accounts, or proxies.
A Reddit discussion asks how teams constrain what coding agents can access and do inside projects. Agents often inherit broad permissions, including `.env` files with API keys and database credentials, and can install dependencies, call APIs, open PRs, or deploy. The poster asks whether teams use short-lived credentials, scoped service accounts, or proxies.
31 engagement1 sourceredditMon, Sep 14, 2026, 02:36 PM