mr-wolv/skill_playground_bakr101_2026/change-risk-assessment
Claude Code and agent skill by mr-wolv/skill_playground_bakr101_2026
- community
- ext:[{"type":"hook","description":"Pre/post lifecycle hooks"}]
Checking that the SKILL.md file is still in the repository.
Assess the delivery risk of a code, config, schema, or infrastructure change and recommend safeguards such as canaries, approvals, rollback plans, and preflight checks.
Install
npx vskill@latest install mr-wolv/skill_playground_bakr101_2026/change-risk-assessmentSemi-extensibleMentions customization, but needs skill-specific knowledge. Read the extensible skills standard
SKILL.md
The same SKILL.md was published earlier as mr-wolv/skill_playground_bakr101_2026/change-risk-assessment.
Change Risk Assessment
Use this skill before shipping a meaningful change when you need to understand blast radius, failure modes, and what safeguards should be in place.
This skill is about deciding how risky a change is and what protections it needs before rollout. It is not just deployment planning; it is preflight judgment.
Use when
- evaluating whether a change is safe to ship
- deciding whether a rollout needs canarying, feature flags, or approvals
- assessing config, migration, infra, auth, or data changes before release
- preparing a preflight checklist for a risky deploy
- comparing rollout options based on risk
Boundary
Use this skill before rollout when the key question is risk: blast radius, reversibility, detectability, failure modes, and which safeguards are justified.
Use shipping-and-launch instead when you already intend to ship and need the practical rollout checklist, staged rollout sequence, monitoring plan, and rollback preparation.
Use operational-readiness-review instead when the question is whether the service or feature is supportable and operable in production as a system, not just whether one change is risky.
Do not use when
- you are already actively responding to a failure — use
incident-response - you need a broad launch checklist and staged rollout plan — use
shipping-and-launch - you need a post-incident analysis — use
postmortemorfailure-analysis
Risk dimensions
Assess the change across these dimensions:
- Blast radius
- how many users, systems, or teams are affected if this goes wrong?
- is this localized, regional, tenant-specific, or platform-wide?
- Reversibility
- can the change be rolled back quickly?
- does rollback restore the previous state cleanly?
- are there irreversible data mutations or contract changes?
- Change surface area
- how many components are changing?
- does the change cross boundaries like app + DB + infra + clients?
- are multiple systems being coordinated at once?
- Operational novelty
- is this a known, repeated workflow or a first-time maneuver?
- are operators familiar with the procedure?
- are tools, scripts, and dashboards already prepared?
- Failure detectability
- would failure be visible quickly?
- are metrics, logs, alerts, and health checks in place?
- could the failure be silent or delayed?
- Data and security sensitivity
- does it affect auth, permissions, secrets, money, or regulated data?
- could failure cause corruption, leakage, or privilege escalation?
- Dependency and coordination risk
- does success depend on external services, ordering, timing, or team coordination?
- are there hidden preconditions?
Risk rating model
Use a simple working model:
- Low risk — small blast radius, easy rollback, strong observability, familiar procedure
- Medium risk — some cross-system effects or rollback caveats, but manageable safeguards exist
- High risk — broad blast radius, difficult rollback, novel procedure, weak detectability, or sensitive data/security implications
Recommended outputs
For each assessed change, produce:
- risk summary
- top failure modes
- likely blast radius
- rollback confidence
- required safeguards
- go / hold / redesign recommendation
Safeguards to recommend
Depending on risk, recommend some combination of:
- feature flags
- canary rollout
- staged rollout
- manual approval gate
- extra reviewer signoff
- schema expand/contract sequencing
- backup or snapshot before change
- preflight verification checklist
- post-deploy watch window
- rollback rehearsal
- freeze window or off-peak timing
Preflight checklist structure
- What is changing?
- What can fail?
- Who/what would be affected?
- How would we know quickly?
- How do we stop or roll back safely?
- What safeguards are mandatory before shipping?
- Who must approve or be on standby?
Warning signs
Escalate risk if you see:
- irreversible migrations
- auth or permissions changes
- coordinated deploys across multiple systems
- unknown dependency behavior
- poor monitoring coverage
- incomplete rollback path
- changes landing before weekends, holidays, or low staffing windows
- "we'll just fix it in prod if needed"
Typical trigger phrases
- "how risky is this deploy"
- "what could break if we ship this"
- "do a preflight risk review"
- "what safeguards should we add before rollout"
- "classify the risk of this change"
- "is this safe to deploy"
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/change-risk-assessment/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Extension points
hookPre/post lifecycle hooks
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/mr-wolv/skill_playground_bakr101_2026/change-risk-assessment)