feat(skills): add /passes for multi-pass planning#203
Merged
emal-avala merged 1 commit intomainfrom Apr 23, 2026
Merged
Conversation
New bundled skill that structures complex tasks into four explicit passes so the agent can't shortcut past important steps: Pass 1 Goal + success criteria — what does "done" mean? Pass 2 Approach + trade-offs — which option wins and why? Pass 3 Implement — scoped to what Pass 1/2 said Pass 4 Verify — run the check named in Pass 1 Each pass has a hard constraint: "don't skip, don't merge, don't expand scope". If Pass 4 fails, loop back to Pass 2 — papering over with commentary is explicitly disallowed. Ends with a one-paragraph summary suitable for a PR description.
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
New bundled skill that structures complex tasks into four explicit passes so the agent can't shortcut past important steps.
The four passes
If Pass 4 fails, the skill explicitly tells the agent to loop back to Pass 2 — papering over a failing check with commentary is disallowed.
Ends with a one-paragraph summary suitable for a PR description: what changed, why that approach, how it was verified.
Why
Single-shot planning misses requirements. Forcing the model to separate what done means from how to get there from the implementation itself surfaces assumption violations early. Bundled skill implementation.
Test plan
cargo fmt --all— cleancargo clippy --workspace --all-targets -- -D warnings— cleancargo test -p agent-code-lib --lib skills— 19/19 pass (no regressions in skill-validator tests)/passesin REPL invokes the skill and starts the agent with the four-pass instructions