chore(deps): update dependency pkdindustries/soulshack to v0.89 - closed

[renovate]

Coming soon: The Renovate bot (GitHub App) will be renamed to Mend. PRs from Renovate will soon appear from 'Mend'. Learn more here.

This PR contains the following updates:

Package	Update	Change
pkdindustries/soulshack	minor	v0.74 -> v0.89

---

Release Notes

pkdindustries/soulshack (pkdindustries/soulshack)

v0.89

Compare Source

What's Changed

refactors to use polly package for llm stuff
we get real streaming in return and net -1800 loc for the bot

new:
--thinking to enable reasoning
--showthinkingactions to announce delays due to reasoning after 5s
--showtoolactions to show toolcalls as an irc action

breaking:
--mcptools -> --mcptool
--shelltools -> --shelltool
--irctools -> --irctool

Full Changelog: pkdindustries/soulshack@v0.87...v0.89

v0.87

Compare Source

Resolved an issue where multiple tool calls in a single assistant response would cause "tool_call_id did not have
response messages" errors.

v0.86

Compare Source

New Features

MCP (Model Context Protocol) Server Support

• Added full MCP server integration - Soulshack can now connect to and use MCP servers as tool providers
• Dynamic tool loading - Automatically discovers and loads tools from MCP servers at startup
• Configuration support - New mcpservers configuration block in YAML config files
• Command-line support - Added --mcpservers flag to specify MCP servers directly

Bug Fixes

• Fixed array property handling for all LLM providers (OpenAI, Anthropic, Gemini, Ollama)
  • Array properties now correctly include the required items field
  • Nested schemas are properly converted recursively
  • Default fallbacks for missing schema information
• Anthropic JSON Schema 2020-12 compliance
  • Fixed schema generation to meet strict JSON Schema draft 2020-12 requirements
  • Properly handles empty properties and required fields
  • Ensures all types have default values when not specified
• Enhanced recursive schema conversion
  • Added proper handling for nested objects and arrays
  • Improved enum value conversion across providers
  • Better type inference and default handling

v0.82

Compare Source

Multi-provider LLM support

• Support for Anthropic Claude, Google Gemini, and Ollama alongside OpenAI
• Models now specified with provider prefix: provider/model
  • openai/gpt-4o
  • anthropic/claude-3-7-sonnet
  • gemini/gemini-2.5-flash
  • ollama/llama3.2

What's New

• Anthropic Claude support with tool calling
• Google Gemini integration with tools
• Ollama for local model hosting
• Refactored tool system with provider-specific converters
• Increased default max tokens to 4096

Breaking Changes

• Model names require provider prefix (was: gpt-4o, now: openai/gpt-4o)

Configuration

• New environment variables: SOULSHACK_ANTHROPICKEY, SOULSHACK_GEMINIKEY, SOULSHACK_OLLAMAURL
• Updated example configs with channel settings

---

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).

🚦 Automerge: Enabled.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.

[renovate]

Edited/Blocked Notification

Renovate will not automatically rebase this PR, because it does not recognize the last commit author and assumes somebody else may have edited the PR.

You can manually request rebase by checking the rebase/retry box above.

⚠️ Warning: custom changes will be lost.