Repository navigation
feat(claude): require self-contained sub-agent prompts - #124
Merged
Merged
Conversation
Loading the agent-instructions skill before every delegation cost context for one-shot prompts. State the rules a delegation prompt needs inline, including how to fence untrusted text, and load the skill only for prompts stored in a file.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The user instructions told Claude to load the
writing:agent-instructionsskill before writing any sub-agent prompt. That loaded the whole skill for every one-shot delegation, even though most of its rules cover durable instruction files.This PR replaces that rule with the requirements a delegation prompt actually needs:
The skill still loads for prompts that live in a file, such as agent definitions, workflow scripts, and skill steps, where its broader rules apply.