skills/scope/SKILL.md
Decompose an accepted spec (and its ADR, if one exists) into GitHub issues using the issue-creator tool.
npx skillsauth add lightforgelabsstudio/aide scopeInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Turn an accepted system spec into a set of GitHub issues. Read AGENTS.md if not in context.
Path to the accepted spec in docs/specs/ (or paste contents). If the design produced an ADR, read it too. Optionally: repo override (owner/repo).
Read the spec (and ADR if present) — Extract: the change being made, Invariants affected, Success Criteria, and any explicit scope boundaries.
Identify work units — Break the decision into discrete, independently deliverable chunks. Each chunk becomes one issue.
Write spec file — Create a temporary spec in issue-creator batch format:
## Epic: <title>
Goal: <from the spec's design/change section>
Spec: <path-to-spec> (plus ADR path if one exists)
### Issue: <title>
Goal: <what this chunk delivers>
Success Criteria:
- <measurable criterion>
Non-Goals: <explicit exclusions>
Run issue-creator — Execute:
python .aide/tools/issue-creator/issue-creator.py <spec-file>
Verify — Confirm each issue was created with correct labels and GitHub Issue Type. Output issue URLs.
Cleanup — Delete the temporary spec file.
.aide/docs/agents/issue-creator-ref.md.aide/tools/issue-creator/README.mdtesting
Skeptical gap analysis for issues, ADRs, plans, and docs. Use when the user wants a second-pass check for missing decisions, contradictions, edge cases, hidden dependencies, scope drift, or readiness before implementation.
tools
Sync branch with remote and confirm a clean end-of-session state.
testing
Create or update an AIDE skill in .aide/skills with valid frontmatter and a concise workflow.
documentation
Apply findings from a /findings review to a spec or ADR and write a response file