Unverified50% confidenceFactExact time
v2.1.294 fixes the misclassification behavior of instruction-based hooks, enabling them to correctly block according to their literal intent.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/8/2026
First Seen
Valid until: 1/6/2027
Sources
Related Entities
Related Claims
UnverifiedIn previous versions, when prompt-type and agent-type hooks were written as natural language instructions, instructions that were supposed to act as blockers instead allowed operations that should have been blocked.68% similarUnverifiedv2.1.294 optimizes the decision logic of prompt hooks on Stop and SubagentStop events, making Claude less likely to stop a task prematurely.67% similarUnverifiedThe Stop Hook can intercept Claude's termination and inject a prompt as a new instruction to continue execution, which is the key to making the entire automated review system work.62% similarUnverifiedAgentWall provides a rollback hook mechanism that executes registered compensating actions in reverse order when a session fails, a design inspired by the Saga pattern.62% similarUnverifiedThe PreToolUse hook can block tool calls by returning exit code 2, thereby intercepting dangerous operations.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/987090API
curl https://kongchang.com/api/v1/knowledge/claims/987090MCP
get_claim(id=987090)