trellis-session-insight

To retrieve past AI conversation history, search local Trellis session logs to locate prior discussions, decisions, and fixes. The Skill extracts relevant context slices using phase-aware boundaries, enabling cross-session recall without restating previous information.

You can resume debugging a familiar bug by searching Trellis mem for the earlier discussion. The Skill locates the relevant session and reconstructs the reasoning behind the previous fix, allowing you to recover the context you need to continue.

The best way to recall decisions from previous AI brainstorming sessions is to search the local conversation history. The Skill performs retrospective review by pulling out implementation history to summarize what changed, what was decided, and what patterns repeat.

Session recall works with the OpenCode adapter but accounts for its limitation by applying phase-aware boundaries during context slicing. This ensures relevant session extraction is accurate despite the adapter constraint.

To finish work from a prior session, ask the assistant to search Trellis mem for the task. The Skill locates the most relevant session and reconstructs the necessary context, enabling task continuity after a gap.

Cross-session conversation recall is limited by the OpenCode adapter limitation, requiring phase-aware boundaries for accurate context slicing. The Skill depends entirely on local Trellis session logs for extracting prior decisions and debugging history.