Document the skills the plugin carries - #18
Merged
Merged
Conversation
The README never named the internal skills — not the prompting one this fork just replaced, nor the two beside it. They are not user-invocable, but they decide how a request reaches Codex and how its answer comes back, so their absence left the reader with no way to know why a rescue behaves as it does. The prompting skill gets the notes that matter in practice here: its launch lines use full model slugs because this fork maps only `spark`, `xhigh` is accepted despite the skill's advice to stay at or below `high`, and a test pins both against the plugin's own validators. The result-handling row names the rule most likely to surprise someone: after presenting review findings Claude stops and asks, rather than applying them. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UXfvnjSC72HsM6EEPVt2Tg
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 README never named the internal skills — not the prompting one just replaced in #17, nor the two beside it. They are not user-invocable, but they decide how a request reaches Codex and how its answer comes back, so their absence left no way to know why a rescue behaves the way it does.
A new Skills The Plugin Carries section under Codex Integration lists all three:
codex-cli-runtimetaskonce, return stdout unchanged, and the list of things the subagent may not dogpt-6-promptingcodex-result-handlingThree notes on the prompting skill, because they are the ones that bite:
gpt-6-luna,gpt-6-sol,gpt-6-astra); this fork maps onlyspark, and every other--modelvalue is forwarded verbatim--effort xhighis accepted — the skill's advice to stay athighor below is about cost, not a limit;maxandultraare not accepted--model/--effortvalues against the plugin's own validators, so the guidance cannot drift into naming something that would be rejected or passed through as a literal wordVerified each claim against the files rather than from memory: the
codex-result-handlingrow was rewritten after reading it, because the first draft called it "verbatim, without a Claude-side summary" and missed its most user-visible rule — the no-auto-fix stop.355/355,
tscclean. Docs only; no behavior change.🤖 Generated with Claude Code
https://claude.ai/code/session_01UXfvnjSC72HsM6EEPVt2Tg
Generated by Claude Code