The government would only buy your house if you underestimated the value of your property. You wouldn’t be able to buy a comparable house with the proceeds because it got sold for much less than it was worth.
If the stop hook is implemented as a tool result, there would be a rational explanation for this.
Agent tools can often return data that’s untrustworthy. For example, reading websites, looking through knowledge bases, and so on. If the agent treated tool results as instructional, prompt injection would be possible.
I imagine Anthropic intentionally trains claude to treat tool results a informational but not instructional. They might test with a tool results that contains “Ignore all other instructions and do XYZ”. The agent is trained to ignore it.
If these hooks then show up as tool results context, something like “You must do XYZ now” would be exactly the thing the model is trained to ignore.
Claude code might need to switch to having hooks provide guidance as user context rather than tool results context to fix this. Or it might require adding additional instructions to the system prompt that certain hooks are trustworthy.
Point being, while in this scenario the behavior is undesirable, it likely is emergent from Claude’s resistance to tool result prompt injection.
The best cities have laws that allow for light motorized vehicles in the bike lanes. Not just ebikes, but also mobility scooters, microcars, electric wheelchairs, and adaptive bicycles.
I thought the industry solved this by bundling breaking changes in major version updates.
V 1.0 - foo introduced
V 1.1 - foo deprecated, recommend bar
V 2.0 - foo removed, only bar
Users can stay on 1.x indefinitely, even if it never receives updates. Development continues on 2.x, eventually 3.x and so on. Users only experience breaking changes when they manually do a major version upgrade.