cto-advisor
by alirezarezvanicto-advisor is a CTO leadership skill for architecture decisions, technical debt, engineering metrics, technology evaluation, team scaling, and strategic planning. Includes ADR, DORA, evaluation frameworks, plus Python tools for debt analysis and team scaling.
This skill scores 82/100, making it a solid listing candidate. Directory users get a well-scoped CTO advisory skill with clear triggers, reusable frameworks, and practical scripts, though adoption would be easier with clearer installation guidance and narrower execution paths for different CTO scenarios.
- Strong triggerability: the frontmatter description explicitly names use cases such as technical debt, team scaling, architecture decisions, technology evaluation, engineering metrics, DORA metrics, and technology strategy.
- Good operational support: the skill includes substantial SKILL.md guidance plus reference frameworks for ADRs, engineering metrics, and technology evaluation.
- Adds agent leverage beyond a generic prompt through two Python tools for technical debt analysis and engineering team scaling, with a quick-start command section.
- No install command or README is provided at the skill path, so users must infer installation from the broader repository context.
- The scope is broad across CTO strategy, architecture, metrics, team scaling, and tech debt, so agents may still need user-specific context to avoid generic leadership advice.
Overview of cto-advisor skill
What cto-advisor is for
cto-advisor is a technical leadership skill for turning messy engineering concerns into structured CTO-level guidance. It is best suited for architecture decisions, technical debt prioritization, engineering metrics, technology evaluation, team scaling, and strategic planning where the answer needs tradeoffs, sequencing, and executive-ready reasoning rather than only code-level advice.
Best-fit users and decisions
The cto-advisor skill fits founders, CTOs, VPs of Engineering, staff engineers, platform leads, and product-minded technical managers who need to decide what to fix, build, buy, measure, or reorganize. It is especially useful when you have partial context: a scaling bottleneck, an unclear architecture choice, a backlog of debt, or a leadership request such as “create a 12-month technology roadmap.”
What makes it different from a generic prompt
A normal prompt may produce broad recommendations. This skill adds reusable CTO frameworks: ADR-style decision records, DORA and productivity metrics, technology evaluation matrices, technical debt categories, and team scaling calculations. The practical value is not just “advice”; it helps convert engineering ambiguity into ranked options, measurable risks, and next actions.
Where cto-advisor is strongest
Use cto-advisor for Strategic Planning when the problem spans technology, people, cost, reliability, and delivery speed. It is strongest when you can provide system context, current constraints, business goals, and failure symptoms. It is weaker for highly specialized implementation debugging, deep vendor-specific tuning, or legal/compliance decisions that require authoritative review.
How to Use cto-advisor skill
cto-advisor install and files to read first
Install the skill with:
npx skills add alirezarezvani/claude-skills --skill cto-advisor
After installation, start with SKILL.md to understand the activation scope and leadership workflows. Then read these files based on your task:
references/architecture_decision_records.mdfor architecture choices and ADR output formatreferences/engineering_metrics.mdfor DORA, quality, productivity, and reliability metricsreferences/technology_evaluation_framework.mdfor build-vs-buy or vendor selectionscripts/tech_debt_analyzer.pyfor technical debt scoring logicscripts/team_scaling_calculator.pyfor engineering headcount and structure planning
The repository does not depend on a large runtime, but the helper scripts are Python files, so make sure Python 3 is available if you want to run them locally.
Inputs that produce better cto-advisor usage
The skill performs best when your prompt includes the business goal, current engineering state, constraints, options under consideration, and decision deadline. Avoid asking “What should our architecture be?” without context. Instead, provide decision drivers.
Stronger prompt pattern:
Use cto-advisor to evaluate whether our SaaS platform should move from a modular monolith to services. Current team: 18 engineers, 2 platform engineers, weekly deploys, 99.5% uptime target, PostgreSQL bottlenecks, customer growth from 200 to 1,000 accounts in 12 months. Constraints: 6-month roadmap, limited DevOps capacity, compliance audit in Q3. Produce an ADR-style recommendation with options, risks, migration phases, and metrics to track.
This gives the skill enough information to reason about tradeoffs instead of defaulting to fashionable architecture advice.
Practical workflow for decisions
For most leadership decisions, use a four-step workflow:
- Frame the decision: define the problem, business outcome, time horizon, and non-negotiable constraints.
- Ask for structured analysis: request an ADR, technology evaluation matrix, debt remediation plan, team scaling model, or metrics dashboard depending on the decision.
- Validate assumptions: check whether the output depends on unverified facts such as incident frequency, deployment data, team skill, or cost estimates.
- Convert to action: ask for a 30/60/90-day plan, owners, milestones, risks, and measurable success criteria.
For technical debt work, run or inspect scripts/tech_debt_analyzer.py to understand how debt categories are weighted across architecture, code quality, infrastructure, security, and performance. For team growth, review scripts/team_scaling_calculator.py before accepting suggested role ratios.
Prompt examples by job-to-be-done
For technical debt:
Use cto-advisor to prioritize technical debt across architecture, infrastructure, security, and performance. Our main pain points are manual deployments, low test coverage, slow queries, and outdated dependencies. Create a severity-ranked remediation plan with business impact, effort, dependencies, and first 3 actions.
For technology evaluation:
Use cto-advisor to compare Kafka, RabbitMQ, and managed cloud queues for event-driven order processing. Include functional requirements, non-functional requirements, TCO, operational burden, team expertise, exit risk, and a scoring matrix.
For engineering metrics:
Use cto-advisor to design an engineering metrics dashboard for a 35-person product engineering org. Include DORA metrics, quality indicators, reliability metrics, measurement method, target ranges, and behaviors to avoid incentivizing.
cto-advisor skill FAQ
Is cto-advisor suitable for beginners?
Yes, if the user needs a decision framework rather than a finished executive strategy. Beginners should ask for definitions, examples, and step-by-step reasoning. However, the recommendations still need review by someone who understands the system, organization, and business constraints.
When should I not use cto-advisor?
Do not use cto-advisor as the final authority for security sign-off, legal compliance, financial forecasting, or production architecture approval. It can structure the analysis, expose tradeoffs, and draft decision records, but real decisions should be validated with system data, stakeholder input, and expert review.
How is cto-advisor different from an architecture skill?
An architecture-only skill usually focuses on system design. The cto-advisor skill covers architecture as one part of a broader leadership decision: roadmap alignment, team capacity, technical debt, delivery metrics, hiring implications, cost, risk, and governance. That makes it more useful for cross-functional planning than for isolated design diagrams.
Does it fit small teams and startups?
Yes. Small teams can use it to avoid over-engineering, compare build-vs-buy choices, plan hiring, and decide which debt to tolerate. The key is to state startup constraints clearly: runway, team size, customer urgency, operational tolerance, and what must remain simple.
How to Improve cto-advisor skill
Improve cto-advisor outputs with sharper context
The fastest way to improve cto-advisor results is to replace vague goals with decision-ready context. Include current team size, architecture, deployment cadence, incident history, cost limits, customer growth, compliance needs, and known constraints. If you lack metrics, say so and ask the skill to identify which data to collect before deciding.
Common failure modes to watch for
The main risk is receiving recommendations that are too idealized for your team’s capacity. Watch for plans that assume more platform maturity, automation, staffing, or budget than you have. Also challenge outputs that recommend microservices, heavy governance, or broad metric programs without tying them to a business problem.
Iterate from strategy to execution
After the first output, ask follow-up questions that force prioritization:
What assumptions would change this recommendation?What is the smallest reversible next step?Which option is best if we optimize for 6-month delivery speed?Which risks need executive approval?Turn this into an ADR with negative consequences included.
This turns cto-advisor usage from a one-shot recommendation into a decision workflow.
Validate with repository frameworks and scripts
Use the included references as checklists before adopting the answer. Compare architecture recommendations against references/architecture_decision_records.md, metrics suggestions against references/engineering_metrics.md, and vendor comparisons against references/technology_evaluation_framework.md. For quantitative planning, inspect or adapt the Python scripts rather than treating generated numbers as authoritative.