Repository navigation
Conversation
The claude runner passed sub-agents as an array of
{name, description, instructions, triggers}. Claude Code expects an
object keyed by agent name, where each entry needs `description` and
`prompt`. Since v2.1.242 it validates this at startup and exits with
"Invalid --agents configuration: Invalid input: expected record,
received array", so `opengap run -a claude` failed for any agent with
an `agents:` block, including examples/full and
examples/nvidia-deep-researcher.
Key each sub-agent by name, rename `instructions` to `prompt`, and fold
delegation triggers into the description, since Claude Code has no
`triggers` field.
Run runWithClaude against a fake claude binary that records its argv, and check that sub-agents are passed as an object keyed by name with `description` and `prompt`, carrying the sub-agent's SOUL.md and its delegation triggers, and that --agents is omitted when the manifest declares no sub-agents. These are the first tests for a runner. They run with `npm test` after `npm run build`, and are skipped on Windows, where the shebang-based fake binary can't run.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
The claude runner passed sub-agents as an array of {name, description, instructions, triggers}. Claude Code expects an object keyed by agent name, where each entry needs
descriptionandprompt.Key each sub-agent by name, rename
instructionstoprompt, and fold delegation triggers into the description, since Claude Code has notriggersfield.Why
Since v2.1.242 it validates this at startup and exits with "Invalid --agents configuration: Invalid input: expected record, received array", so
opengap run -a claudefailed for any agent with anagents:block. https://code.claude.com/docs/en/errors#invalid-agents-configurationHow Tested
npm run buildpassesgitagent validatepasses on example agentsexamples/fullusingrun -d examples/full -a claude -p "hello"Checklist
Closes #106