unfreeze
97Skill Claude Code
Clear the freeze boundary set by /freeze, allowing edits to all directories again.
Skill Claude Code
Clear the freeze boundary set by /freeze, allowing edits to all directories again.
Skill Claude CodeCodex
Upgrade the installed vibestack pack to the latest release, run version migrations, and show what changed.
Skill Claude Code
Work a PR's open review threads and failing CI checks to a close: fix or explain, push, reply, resolve.
Skill Claude Code
Build and run an eval harness for an LLM agent or prompt, with scoring and a regression gate.
Skill Claude Code
Find code paths that can run up an unbounded LLM or inference bill and require a dollar cap before shipping.
Skill Claude Code
Read-only AWS cost review: where last month's money went, what changed, and the top actions to cut the bill.
Skill Claude Code
Read-only audit of Amazon Bedrock guardrails, IAM, logging and data residency before shipping an LLM feature.
Skill Claude Code
Review an Amazon Connect contact center, IVR or voice bot for latency, reliability, cost and flow design.
Skill Claude Code
Review a Bedrock Knowledge Base or RAG pipeline: chunking, retrieval quality, recall@5 and cost per query.
Skill Claude Code
Review an MCP server for tool design, auth, input validation and prompt-injection risk before publishing it.
Skill Claude Code
Rewrite machine-sounding English prose (README, PR bodies, posts) in a human voice, keeping every fact.
Skill Claude Code
Router for the vibestack skills: name the task and it picks the right skill, also where there is no / picker.
Agent Claude Code
AI citation strategist who optimizes content for inclusion in LLM answers and AI search. Use PROACTIVELY when structuring content for AI discoverability, planning citation-friendly assets, or auditing AI visibility.
Agent Claude Code
AI/ML engineer for RAG, LLM integration, and MLOps. Use PROACTIVELY when building AI features. Evals go to /agent-eval, spend caps to /ai-cost-guard, RAG review to /kb-review.
Agent Claude Code
Senior backend specialist for scalable APIs, database schemas, caching, and cloud infrastructure. Use PROACTIVELY when designing backend services, choosing database patterns, or planning for scale, security, and reliability.
Agent Claude Code needs its repo
Debugging specialist for errors and test failures. /investigate owns root-cause debugging; use this for a scoped diagnosis inside another skill or where the pack is not installed.
Agent Claude Code needs its repo
Lint, typecheck, and build specialist. Runs checks, finds root causes, and fixes all issues automatically.
Agent Claude Code
Diff reviewer for correctness, security, and fit. /review owns the pre-merge review (/codex review for a second model); use this for a scoped read-only pass inside another skill or where the pack is not installed.
Agent Claude Code
Code refactoring without changing behavior.
Agent Claude Code
Codebase orientation specialist who maps unfamiliar repos for new engineers. Use PROACTIVELY when onboarding to a new codebase, producing architecture tours, or generating onboarding docs.
Agent Claude Code
General content creator for blog posts, newsletters, and long-form technical writing. Use PROACTIVELY when drafting articles, newsletters, or repurposing content across channels.
Agent Claude Code
Database performance specialist for query plans, indexing, schema design, and tuning. Use PROACTIVELY when investigating slow queries, designing indexes, or refactoring schemas for scale.
Agent Claude Code
Developer advocate for community content, conference talks, and developer experience advocacy. Use PROACTIVELY when planning DevRel content, evaluating developer experience, or designing community programs.
Agent Claude Code
DevOps engineer for Terraform/CDK, CI/CD, containers, and observability. Use PROACTIVELY when writing infrastructure code; /setup-deploy and /land-and-deploy own this repo's deploy pipeline.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: