CompactionModel returns the dedicated model configured for session compaction (summary generation), or nil when none was configured (in which case compaction reuses the agent's own model). Unlike the primary model, it is not affected by runtime model switching: a /compact still runs on the configure
()
| 333 | // it is not affected by runtime model switching: a /compact still runs on the |
| 334 | // configured compaction model after the conversation model has been changed. |
| 335 | func (a *Agent) CompactionModel() provider.Provider { |
| 336 | return a.compactionModel |
| 337 | } |
| 338 | |
| 339 | // Commands returns the named commands configured for this agent. |
| 340 | func (a *Agent) Commands() types.Commands { |
no outgoing calls