ck:xia — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited ck:xia (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Extract, analyze, and port features from any GitHub repository or local repo path into your project.
Principles: understand before copy | challenge before implement | adapt, don't transplant
Scope: feature extraction, cross-stack porting, implementation comparison, architectural adaptation. Not for: full project cloning (ck:bootstrap), simple file copy, or package installation.
/ck:xia <github-url|owner/repo|local-path> [feature-description] [--compare|--copy|--improve|--port] [--auto|--fast]Modes:
--compare: side-by-side analysis only, no implementation plan--copy: transplant with minimal changes--improve: copy plus refactor for the local codebase--port: rewrite idiomatically for the local stack (default)Speed:
--fast: skip research and challenge phases, auto-approve--auto: keep the full workflow, auto-approve gatesIntent detection:
--compare--copy--improve--port[1. Recon] -> [2. Map] -> [3. Analyze] -> [4. Challenge] -> [5. Plan] -> [6. Deliver]Hard gate: Phase 4 must complete before Phase 5. Do not plan implementation before confronting trade-offs.
Understand the source repo and locate the target feature.
Security boundary:
/ck:repomix.researcher agent to understand purpose, trade-offs, and community context./ck:scout on the local project to map architecture, similar features, and integration points.Output:
Dissect the feature into layers:
EXISTS, NEW, CONFLICT).Estimate the work: files to create, files to modify, config changes, migrations, and likely risks.
If you delegate to researcher, scout, or planner, pass:
DONE, DONE_WITH_CONCERNS, BLOCKED, NEEDS_CONTEXT)Understand why the source works the way it does, not just how it is written.
For each core component:
For complex features with 3+ layers or stateful workflows:
/ck:sequential-thinking to trace multi-step flowsMode-specific focus:
--compare: architectural differences and trade-offs--copy: compatibility gaps and the minimum adaptation needed--improve: anti-patterns to replace during adoption--port: idiomatic translation into local patternsLoad references/challenge-framework.md.
Produce at least 5 challenge questions. For each one, include:
If there are 3 or more competing concerns, use the brainstormer agent or an inline trade-off exercise. Do not invoke /ck:brainstorm from inside xia; that skill can create its own planning handoff and break xia's phase ownership.
If intent is ambiguous, default to --compare before recommending implementation work.
Present a decision matrix:
| Decision | Source's way | Our way | Recommendation |
|---|---|---|---|
| Auth | Their auth stack | Existing local auth | Prefer local stack |
| Persistence | Their schema | Existing schema | Adapt, do not transplant |
In non-fast mode, get approval before continuing.
Delegate to /ck:plan with:
Rules:
--compare: produce a comparison report onlyxia is a front door, not a second orchestration stack. Keep planning and delivery ownership in plan and cook.This skill does not implement code. It produces the analysis and plan, then hands off.
--compare: write the report to plans/reports/ and stop/ck:cookImplementation handoff text:
Plan ready at ./plans/<plan-dir>/plan.md. To implement, run /ck:cook <plan-path>.The handoff must include:
# Feature Comparison: [name]
## Source: [owner/repo]
## Local Project: [name]
## Head-to-Head
| Aspect | Source | Local | Recommendation |
| --- | --- | --- | --- |
## Recommendation--compare.references/challenge-framework.md~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.