Build(deps-dev): Bump amrita-core from 0.13.5 to 0.13.7

[dependabot]

Bumps amrita-core from 0.13.5 to 0.13.7.

Release notes

Sourced from amrita-core's releases.

V0.13.7

What's Changed

• Use active model preset for memory summarization by @​JohnRichard4096 in AmritaBot/AmritaCore#158

Full Changelog: AmritaBot/AmritaCore@0.13.6...0.13.7

V0.13.6

We are pleased to announce the release of AmritaCore v0.13.6. This update brings significant improvements to the preset-fallback mechanism, better error handling, and a clearer, more explicit default‑preset behaviour. Please review the breaking changes below before upgrading.

✨ New Features

• Specialised fallback event contexts
  The former FallbackContext is now a base class, with three concrete subclasses that correspond to the specific gateway call that failed:

  • CompletionFallbackContext – fired when call_completion fails.
  • ToolsFallbackContext – fired when tools_caller fails (also exposes the tool schemas of the failed call).
  • EmbeddingFallbackContext – fired when call_embedding fails.

  This allows matchers registered via @on_preset_fallback() to distinguish the kind of failure and react accordingly (e.g., by inspecting the tools field, or by using the specific payload type of the failed call).

• Gateway‑level fallback loop
  The preset‑fallback retry logic has been moved from the LLM_COMPLETION workflow node into the gateway functions (call_completion, tools_caller, call_embedding). This makes fallback handling consistent across all LLM‑related gateway calls and decouples it from the workflow layer.

⚠️ Breaking Changes

• PresetManager.get_default_preset() and MultiPresetManager.get_default_preset() no longer fall back to a random preset when no default is set.
  Instead, they now raise RuntimeError immediately with the message:
  "Default preset not set, use set_default_preset() first."
  This “fail‑fast” behaviour surfaces misconfiguration early and replaces the previous silent (and often unintended) random fallback.

  Migration: If your code relied on the automatic random fallback, you must now explicitly call set_default_preset() before using get_default_preset().

• The FallbackContext class is now abstract in nature – it serves as a base class only.
  If you were manually instantiating FallbackContext (e.g., for testing), you should now use the appropriate subclass (CompletionFallbackContext, ToolsFallbackContext, or EmbeddingFallbackContext) instead.

🔧 Improvements

• Documentation overhaul
  All API reference pages for FallbackContext, PresetManager, and MultiPresetManager have been updated to reflect the new class hierarchy, the fail‑fast default behaviour, and the changed semantics of get_default_preset().

• Enhanced event typing
  The context field in FallbackContext now uses a union type (SendMessageWrap | CONTENT_LIST_TYPE | Sequence[str]) that reflects the actual payload carried by each subclass, improving type safety.

• Internal code clarity
  The fallback logic has been centralised into helper functions (_fire_fallback and _with_preset_fallback), making the gateway functions cleaner and more maintainable.

• New test suite
  Added test_fallback_gateway.py to comprehensively cover the new gateway‑level fallback behaviour, including event firing, preset swapping, and exhaustion of retries.

PRs

... (truncated)

Commits

• 1ddd98f Use active model preset for memory summarization (#158)
• 3849551 Move preset fallback to gateways and fail fast on missing default (#157)
• See full diff in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)