CLAUDE.md: proactive gap-filing directive + vendor write-eigenscript skill#679
Merged
Merged
Conversation
…skill Two contributor-onramp changes so an external AI building in EigenScript has the same guidance the maintainer's agents do. 1. A standing, imperative directive near the top of CLAUDE.md: when you hit a language gap, rough edge, or bug, file a GitHub issue proactively (or fold the fix into your PR) rather than working around it silently. The forcing-function model was already described in the Ecosystem section, but only descriptively — this makes it an instruction the agent acts on unasked. 2. Vendor the write-eigenscript skill into .claude/skills/ so Claude Code auto-loads it for any contributor in the repo. It encodes the silent-wrong traps (of-call spread rule, outward-mutable scope / when to use local, soft-keyword pitfalls, observer/temporal idioms) that are otherwise absent outside the maintainer's own machine. Scrubbed for public consumption: examples that referenced private/unshared repos were genericized; all language guidance is verbatim. This is the missing half of the on-ramp — an external AI given write-eigenscript writes correct .eigs, and given the gap directive surfaces its gaps upstream, so adoption feeds back as contribution (the forcing-function model, externalized). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
Two contributor-onramp changes so an external AI building in EigenScript gets the same guidance the maintainer's agents do.
1. Proactive gap-filing directive — a standing, imperative section near the top of CLAUDE.md: hit a gap, rough edge, or bug → file a GitHub issue (or fold the fix into your PR), don't work around it silently, do it without being asked. The forcing-function model was already in the Ecosystem section but only described; this makes it an instruction the agent acts on unprompted.
2. Vendor
write-eigenscriptinto.claude/skills/so Claude Code auto-loads it for any contributor in the repo. It encodes the silent-wrong traps — theof-call spread rule, outward-mutable scope / when to uselocal, soft-keyword pitfalls, observer/temporal idioms — that are otherwise unavailable outside the maintainer's own machine.Public-consumption scrub
The skill referenced private/unshared repos in three examples (an EigenOS desktop helper, a named consumer's piano-roll UI). Those were genericized so an external contributor isn't pointed at code they can't see. All language guidance is verbatim — only the illustrative references changed.
Why
This is the missing half of the on-ramp: an external AI given
write-eigenscriptwrites correct.eigs, and given the gap directive surfaces its gaps upstream — so adoption feeds back as contribution (the forcing-function model, externalized).Docs + skill only, no code.
🤖 Generated with Claude Code