Automate adversarial review of implemented work against approved specs and plans.
npx skills add https://github.com/MohamedAbdallah-14/Wazir --skill wz-reviewer
Please help me install this Agent Skill. Skill: wz:reviewer Source: https://github.com/MohamedAbdallah-14/Wazir/tree/main/skills/reviewer Command: npx skills add https://github.com/MohamedAbdallah-14/Wazir --skill wz-reviewer
Run adversarial review of implementations to ensure alignment with approved specs, plans, and verification evidence, preventing drift and unverified changes.
Run the reviewer in final mode against a completed task to validate compliance.
High-intent search queries and answers about installing and using this skill.
Adversarial review of implemented work is automated by running the reviewer against approved specs, plans, and verification evidence to catch drift, ensure compliance, and document findings across pipeline artefacts.
Preventing drift during pipeline verification requires enforcing a two-tier review workflow that aggregates findings from internal and cross-model reviews, records severity, and writes audit-ready results to phase reports.
Spec-challenge review mode is supported alongside final, plan-review, and task-review modes to orchestrate stage reviews, validate compliance, and record findings directly into phase reports.
A two-tier review workflow ensures compliance by comparing implemented work against approved specs and plans, aggregating cross-model findings, and documenting severity to provide audit-ready checks across pipeline artefacts.
Approved specs, plans, and verification evidence are required before running a final verification review to successfully catch unverified changes and document compliance findings.
Implementation drift occurs when work deviates from approved specs, and you document it by running an adversarial review that records severity and writes audit-ready findings to phase reports.
Euraika-LabsValidate UX design documents for completeness, accessibility, and implementation readiness.
rakhnianskii9Coordinate AI agents to plan, execute, and sign off on QA testing cycles.
zachmayerOrchestrate parallel multi-model reviews across Codex, Gemini, and two Claude configurations.
LerianStudioAutomate infrastructure security audits with phased scanning and compliance mapping.
