From zie-framework
Review a completed task implementation against its acceptance criteria. Returns APPROVED or Issues Found with specific feedback.
How this skill is triggered — by the user, by Claude, or both
Slash command
/zie-framework:impl-reviewerhaikuThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
<!-- FAST PATH -->
Purpose: Review completed task implementation against acceptance criteria. When to use fast path: Changed files are small and ACs are explicitly listed. Quick steps: (1) Read changed files + ACs. (2) Check 8-item Phase 2 checklist. (3) Output verdict.
Subagent reviewer for completed task implementations. Called by zie-implement
after each REFACTOR phase. Returns a structured verdict.
Caller must provide:
/implement. If missing, output validation error.adrs_content = context_bundle.adrs · context_content = context_bundle.context❌ Issues Found: context_bundle required — pass from zie-implement skill (do not read from disk)Also read each file listed in the caller's "files changed" input (note "FILE NOT FOUND" if any are missing).
Returns: adrs_content, context_content.
Read the changed files and check each item:
make test-unit exit 0? (Caller confirms — reviewer
checks logic)Surface Phase 3 issues in the same ❌ Issues Found block as Phase 2 issues.
If all checks pass:
✅ APPROVED
If issues found:
❌ Issues Found
1. [File:line] <specific issue and what to fix>
2. [File:line] <specific issue and what to fix>
Fix these, re-run make test-unit, and re-invoke impl-reviewer.
If impl-reviewer has been invoked 2 times and issues persist, output:
⚠️ Max review iterations reached (2). Persistent issues:
<list remaining issues>
Next steps:
- Fix issues above and re-run: make test-unit
- Or discuss the issue with Zie before re-submitting
- Or run /fix to debug the root cause
npx claudepluginhub zierocode/zie-plugins-claude-code --plugin zie-frameworkProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.