We just shipped agent context audits at Alignbase! Now you can know exactly what your agents knew, every time they acted.
When an agent takes action, Alignbase records the context bundle behind that run: the policies, AGENTS.md entries, Skills, versions, tags, and routing decisions that shaped the session. That gives teams a point-in-time record they can inspect later instead of guessing which instructions were present.
This helps with review, debugging, and governance. If an agent misses a rule or follows stale guidance, you can trace the input path, fix the source context, and prevent the same failure from spreading across the rest of the fleet.
Top comments (2)
This is the real gap. When an agent misbehaves, the hardest part is usually reconstructing what it actually saw, not just what we intended it to see. I'm actually curious, how granular the context bundle gets, does it capture the final resolved prompt after all the routing/merging logic runs, or just the list of sources that were eligible to be included?
It captures everything injected into the agent, but it doesn't store the prompts (yet).