Releases·Tool Release
hallpass: open-source tool checks whether the human user—not the bot account—is allowed to do an agent's action
A developer released hallpass, an open-source authorization layer that checks whether the individual user prompting an agent is actually allowed to perform a requested action before the agent executes it. It addresses the common pattern where agents act through a single over-privileged bot account, letting any user leverage permissions they lack themselves. For practitioners building agents on Jira, GitHub, Slack, or AWS, it offers a self-hosted way to enforce per-user authorization.
18 engagement1 sourceWed, Sep 23, 2026, 08:28 PM
AI Overview
- hallpass (Apache 2.0, self-hosted) checks per-user permissions before an agent acts, returning allow, deny, or unknown (treated as deny) via a POST /check API
- Most agents use one bot account for Jira, GitHub, Slack, and AWS, so anyone who can talk to the agent can do whatever the bot can
- hallpass works with 11 systems today, ships as one binary or Docker, and asks the real system whether the named user is allowed
- Example check: {"user": "[email protected]", "action": "DELETE_ISSUES", "resource": "issue:PAY-123"} returned a deny decision
Entities
Related
Products·launched
Authorization check tool that gates an ops agent's GitOps PR creation by asking GitHub if the user may push
An internal ops agent that reports service health and proposes changes via pull requests to a GitOps repo. The author added a tool that asks GitHub whether the requesting person is allowed to push to that repo before the agent opens a PR, closing a privilege-escalation gap where the bot's token could act for users without push access.
An internal ops agent that reports service health and proposes changes via pull requests to a GitOps repo. The author added a tool that asks GitHub whether the requesting person is allowed to push to that repo before the agent opens a PR, closing a privilege-escalation gap where the bot's token could act for users without push access.
7 engagement1 sourceredditThu, Sep 24, 2026, 07:36 PM
Social·Community
Developers weigh authorization layers for production agent tool calls that modify external systems
A developer asks how teams handle authorization for agent tool calls in production, beyond demos. The core question is where permission logic should live: in a policy layer between agent and tool, in the API itself, and how to pass per-user identity and permissions through tool calls when tools can trigger irreversible actions.
A developer asks how teams handle authorization for agent tool calls in production, beyond demos. The core question is where permission logic should live: in a policy layer between agent and tool, in the API itself, and how to pass per-user identity and permissions through tool calls when tools can trigger irreversible actions.
18 engagement1 sourceredditTue, Sep 15, 2026, 06:02 PM
Products·concept
Control gateway layer between AI agents and their tools for credentials, permissions, and auditing
A control layer that sits between agents and their tools, centralizing OAuth/credentials, tool permissions, policies and approvals, rate limits, and execution logs. The builder is validating demand with agent teams who currently rebuild this logic in their own backends. It targets production agent deployments with dozens of APIs and custom tools rather than 3-5 toy setups.
A control layer that sits between agents and their tools, centralizing OAuth/credentials, tool permissions, policies and approvals, rate limits, and execution logs. The builder is validating demand with agent teams who currently rebuild this logic in their own backends. It targets production agent deployments with dozens of APIs and custom tools rather than 3-5 toy setups.
10 engagement1 sourceredditSun, Sep 13, 2026, 08:46 AM
Products·building
AgentHail: controlled execution layer for AI agents
AgentHail is a middleware layer that sits between an AI agent and business tools, handling authentication, permissions, human approvals, retries, duplicate prevention, logging, and failure recovery. It ensures agents cannot directly execute sensitive actions, addressing safety and reliability concerns for agent builders.
AgentHail is a middleware layer that sits between an AI agent and business tools, handling authentication, permissions, human approvals, retries, duplicate prevention, logging, and failure recovery. It ensures agents cannot directly execute sensitive actions, addressing safety and reliability concerns for agent builders.
1 engagement1 sourceredditTue, Jul 21, 2026, 03:06 PM
Products·launched
Igris Security — governance layer for AI agents with RBAC and access control for tools
Igris Security is building a governance layer for AI agents in production, controlling which agents can call which tools. It uses policy-based access control (deny-by-default RBAC) to prevent over-privileged agents from taking destructive actions. Aimed at teams shipping LLM agents that connect to tools via MCP with shared credentials.
Igris Security is building a governance layer for AI agents in production, controlling which agents can call which tools. It uses policy-based access control (deny-by-default RBAC) to prevent over-privileged agents from taking destructive actions. Aimed at teams shipping LLM agents that connect to tools via MCP with shared credentials.
5 engagement2 sourcesredditFri, Sep 4, 2026, 11:28 AM