docs(sdk): Add guide for parallel tool execution#409
Conversation
Adds documentation for the experimental tool_concurrency_limit feature that allows agents to execute multiple tools concurrently within a single LLM response. - Created sdk/guides/parallel-tool-execution.mdx with: - Overview of parallel tool execution - Configuration guidance - Use cases and best practices - Experimental warning and safety considerations - Ready-to-run example code - Updated docs.json navigation Co-authored-by: openhands <openhands@all-hands.dev>
all-hands-bot
left a comment
There was a problem hiding this comment.
🟡 Acceptable - Documentation is technically accurate against the feature branch, but has usability issues that should be fixed.
bc3a56f to
442b34e
Compare
442b34e to
d59e8d6
Compare
all-hands-bot
left a comment
There was a problem hiding this comment.
🟢 Good taste - Documentation is accurate, pragmatic, and doesn't over-engineer. Verified against feature branch source code.
…rallel-tool-execution # Conflicts: # sdk/guides/parallel-tool-execution.mdx
all-hands-bot
left a comment
There was a problem hiding this comment.
🟡 Acceptable - Documentation is technically accurate against the feature branch, but the focus highlight still doesn't match the comment intent.
Co-authored-by: OpenHands Bot <contact@all-hands.dev>
Co-authored-by: Engel Nyst <engel.nyst@gmail.com>
enyst
left a comment
There was a problem hiding this comment.
If you prefer you can negotiate with the agent, but since I'm here I'll approve it anyway 😅
Thank you for the PTE work!
all-hands-bot
left a comment
There was a problem hiding this comment.
🟡 Acceptable - Documentation accurately reflects the SDK source (verified against main branch), but has a broken link and minor focus issue.
Co-authored-by: OpenHands Bot <contact@all-hands.dev>
@enyst Thanks :-) I just wanted to be sure I was not adding broken links.... |
Summary
Adds documentation for the experimental
tool_concurrency_limitfeature in the OpenHands Agent SDK that allows agents to execute multiple tools concurrently within a single LLM response.Changes
New Documentation
sdk/guides/parallel-tool-execution.mdxwith:tool_concurrency_limitparameterparallel-tool-executionbranchNavigation
docs.jsonto add the new guide afteragent-delegation(related concept)Notes
tool_concurrency_limitdefaults to1(sequential)