The latest release identifiable for anteriorai/symbolic-reasoning is @anteriorai/srg-parser@0.7.0. This repository is a Python library that deals with the problem of agents missing policy rules or applying them incorrectly. It does not leave policy judgment solely to the model's response, but handles the rules as symbolic logic. However, the provided information contains no concrete list of what changed in this release.
Policies are expressed as graphs in the .srg language. The facts of a case are supplied by the agent or an external reasoner, and the library uses them to determine the next question and result while recording the reasoning path. The oracle that the solver queries is the judging entity that answers the facts of the case. This structure focuses on separating the policy file that sets the rules from per-case factual judgment. It can be run in an ordinary workflow or connected to an agent as a tool call, and the tool server is provided via MCP. The repository language is Python and the license is Apache-2.0. The star count recorded in FACTS is 11.
The API is below 1.0 and is described as being at a research stage. The provided information contains no benchmark showing how much it reduces rule omissions or misapplications. Therefore, the design of processing policy rules as a symbolic graph and recording the path is confirmed, but there is not enough basis to judge the magnitude of the actual effect or its stability.