Objective
Add a governed runx skill proposal for a collaboration issue distillation skill that turns a bounded GitHub issue thread into a reusable approval packet and maintainer recap before any publish or build gate proceeds.
Why It Matters
We want issue review to train the operator. Human rationale should become explicit, rebuildable context rather than hidden prompt sprawl. This run should tell us whether skill-lab can produce a useful design for issue-shaped learning and approval work.
Constraints
- proposal only, no implementation code
- output stays under
docs/skill-proposals/
- skill output should be markdown/json artifacts, not direct GitHub mutation
- treat issue comments and GitHub Actions receipts as canonical evidence; derived memory must stay rebuildable
- call out approval boundaries explicitly and keep any publish surface draft-only
- acceptance checks should cover missing rationale, conflicting comments, and stale evidence
Evidence
state/thread-teaching.json is the current derived teaching surface
docs/philosophy.md says repeated work should become explicit skills, not prompt sprawl
docs/operations.md says every lane has an approval mode and memory effect
- collaboration issues are becoming a core review surface for bounded approvals
Objective
Add a governed runx skill proposal for a collaboration issue distillation skill that turns a bounded GitHub issue thread into a reusable approval packet and maintainer recap before any publish or build gate proceeds.
Why It Matters
We want issue review to train the operator. Human rationale should become explicit, rebuildable context rather than hidden prompt sprawl. This run should tell us whether skill-lab can produce a useful design for issue-shaped learning and approval work.
Constraints
docs/skill-proposals/Evidence
state/thread-teaching.jsonis the current derived teaching surfacedocs/philosophy.mdsays repeated work should become explicit skills, not prompt sprawldocs/operations.mdsays every lane has an approval mode and memory effect