Add agent-handoff (hand-off checklist + real cold-start gate)#138
Open
drakosha wants to merge 1 commit into
Open
Add agent-handoff (hand-off checklist + real cold-start gate)#138drakosha wants to merge 1 commit into
drakosha wants to merge 1 commit into
Conversation
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds agent-handoff to Context Delivery & Compaction.
Why it belongs: hand-off prompts (PROMPT.md for a loop, a brief for a fresh session or peer agent) are a context-delivery artifact, and most guidance stops at "write it carefully." This repo contributes two concrete mechanisms: a scaffold whose sections are sections of the prompt itself (not a post-hoc audit), and
gate.sh— piping the draft into a genuinely fresh session that lists its first 3 questions and answers READY/BOUNCE, replacing self-simulation (which fails in both directions) with a real cold-start run. Includes two worked examples, one with a stop rule firing on a risky investigation.Distilled from daily agent-driven engineering practice; published as-is, no benchmarks claimed.