Skip to content
risk-assessment-writer logo

Risk Assessment Writer

risk-assessment-writer

ACTIVATE when the user asks to write, create, draft, or generate a risk assessment, risk entry, risk evaluation, or threat/vulnerability description — or when the user describes a threat, vulnerability, weakness, new business activity, or scenario they want risk-assessed. Covers information secur...

eyesecurity/skills0installs11stars

SKILL.md

Full skill instructions

Risk Assessment Writer

Every new risk gets a structured assessment. No threat goes undocumented.

All output is written entirely in English.

Storage

  • Naming: [YYYY]-[NN] [Short description] (e.g. "2026-101 Insufficient patch management")

Risk assessment record

Each risk page has two main sections: Risk Evaluation and Risk Treatment.

Risk Evaluation

FieldDescriptionValues
AssetWhat is at riskDescriptive phrase or single word (e.g. "Data", "Software", "Organization", "Services", "People", "Hubspot data", "HR payroll SaaS platform")
ThreatWhat could go wrong and why (see Threat Patterns below)1-4 sentences
CategoryWhich security/​quality dimensionConfidentiality, Integrity, Availability, QUALITY (one or more)
LikelihoodHow likely is this to happenL (Low), M (Medium), H (High)
ImpactHow bad if it happensL, M, H
RiskDerived from matrix (see below)L, M, H
OwnerWho is accountableSecurity Officer, CTO, Management, CISO office, DPO, Quality Manager, Marketing Lead, or a domain-specific role
StatusTreatment decisionACCEPT (Green), REDUCE (Yellow), PREVENT (Red), TRANSFER
Date of last reviewWhen last assessedYYYY-MM-DD
CountriesWhere this risk appliesNetherlands, Belgium, Germany, All

Risk Treatment

FieldDescription
Treatment optionACCEPT, REDUCE, PREVENT, or TRANSFER
MeasuresChecklist: [x] = implemented, [ ] = planned. Can also be free text or "N/​A"
Residual riskMust be <= original Risk level
ResponsiblePerson/​role accountable for the measures

Risk matrix (L/​M/​H)

Impact LImpact MImpact H
Likelihood HMHH
Likelihood MLMH
Likelihood LLLM

GRC-Tool numeric scoring (1-3, use only when user requests it)

  • Likelihood: 1 = Unlikely, 2 = Somewhat likely, 3 = Likely
  • Impact: 1 = Low (<EUR 10k), 2 = Medium (EUR 10k-50k), 3 = High (>EUR 50k)
  • Risk score = Likelihood x Impact: 1-2 Green/​Low, 3-5 Yellow/​Medium, 6-7 Red/​High, 8-9 Purple/​Critical

Default to L/​M/​H unless the user explicitly asks for numeric scoring.

Risk appetite

The default financial thresholds for impact scoring are:

  • L: Under EUR 10k
  • M: EUR 10k-50k
  • H: Over EUR 50k

These defaults may not fit every organization. When starting a new risk assessment (or the first time working with a user), ask:

The default financial thresholds for impact are: L = under EUR 10k, M = EUR 10k-50k, H = over EUR 50k. Do these match your risk appetite, or would you like to adjust them?

If the user provides custom thresholds, use those consistently throughout the assessment and the impact questions. If the user confirms the defaults or wants to skip, proceed with the defaults.

Risk lifecycle

1. Identify

When the user describes a new risk scenario, business activity, or threat:

  • Determine the Asset at risk
  • Propose the most fitting Category (Confidentiality, Integrity, Availability, QUALITY)
  • Draft a Threat description using one of the three patterns below

2. Assess -- Guided Likelihood and Impact

When the user is unsure about Likelihood or Impact, walk them through these questions conversationally (not as a dry checklist). If the user already provides L/​M/​H values, skip this step.

Likelihood questions
#QuestionHML
1Has this type of incident happened before?Multiple times in past 2 yearsOnce or twiceNever, here or at similar orgs
2Are there existing controls in place?No controls or very weakSome controls with gapsStrong, tested controls
3How exposed is the asset or process?Public / many users / internet-facingLimited but shared across teams or vendorsVery restricted, isolated
4How complex to materialize?Simple, single mistakeSpecific chain of events neededMultiple simultaneous failures required

Scoring: Majority H = H, majority L = L, otherwise M.

Impact questions
#QuestionHML
1Financial exposure?Over EUR 50kEUR 10k-50kUnder EUR 10k
2Customer or service delivery impact?Significant, broad customer disruptionPartial, some customers or services affectedNo customer impact, internal only
3Regulatory or legal consequences?Fines, sanctions, lawsuits, or license revocationWarning or minor regulatory findingNone
4Reputational damage?Media coverage or significant loss of customer trustInternal or industry-level attentionNone expected
5Recovery time?Weeks, months, or irreversibleDays to a weekHours, quickly recoverable

Scoring: Count H, M, and L answers. Majority H = H, majority L = L, otherwise M.

Note: The financial thresholds in Q1 follow the risk appetite defaults above. If the user has provided custom thresholds, substitute them here.

Presenting the recommendation

After evaluating, present:

Based on your answers:

  • Likelihood: [L/​M/​H] -- [1-sentence reasoning]
  • Impact: [L/​M/​H] -- [1-sentence reasoning]

This gives a Risk level of [L/​M/​H] per the risk matrix.

Do you agree, or would you like to adjust?

3. Treat

Determine treatment together with the user:

OptionWhen to useMeasure pattern
REDUCERisk is too high, controls can lower it[ ] for planned, [x] for already implemented
ACCEPTRisk is within appetite, or already reduced[x] for all implemented, or "N/​A"
PREVENTRisk must be avoided entirelyDescribe how the activity/​exposure is eliminated
TRANSFERRisk shifted to third party (insurance, contract)Describe the transfer mechanism

4. Document

Generate the full risk assessment in the standard two-table format (Risk Evaluation + Risk Treatment). Suggest a page title following the [YYYY]-[NN] [Short description] convention.

5. Review

  • Risks should be reviewed at least annually
  • High risks: review every 6 months
  • When the business context changes (new country, new vendor, new regulation): re-assess

Threat description patterns

Pattern A -- Formal formula:

As a result of [root cause 1] OR [root cause 2], there is a chance that [undesirable event], resulting in [business impact].

Pattern B -- Narrative style: 2-4 sentences describing the context, why the risk exists, and the potential consequence.

Pattern C -- Short direct causal:

Due to [cause], it is possible that [event]. This could lead to [impact].

Root cause library

DomainExamples
ProceduralLack of policies, insufficient authorization policies, absence of guidelines, incomplete processes, missing acceptance criteria
TechnicalInsufficient monitoring, unsecured communication, lack of logging, poor change management, insufficient segmentation, missing patch management
OrganizationalInsufficient segregation of duties, lack of knowledge, missing management involvement, poor vendor management, insufficient awareness training
PhysicalFlammable materials, unsafe premises, shared offices, multiple locations, utility failure
Third-partyUS-based vendor (data residency), vendor with environment access, lack of legal/​technical separation, missing vendor logging
HR/​PeopleDeparting employees, dual roles, undesired behavior, language proficiency gaps, single point of failure

Impact library

DomainExamples
ConfidentialityData accessible to unauthorized persons, unauthorized distribution, competitive position harmed, data leaked to third party
IntegrityData processed incorrectly, no segregation of duties, unsafe code deployed
AvailabilityBusiness continuity at risk, long recovery times, services not delivered
ComplianceSanctions, regulatory non-compliance, fines, privacy violations
FinancialRevenue loss, material damage, liability, claims
ReputationMedia attention, loss of customer trust
QualityService quality degraded, projects not delivering results

Measure library

CategoryExample measures
Access controlAccess policies, user provisioning, access rights revocation, password management, secure login, SSO, IP allow listing, VPN, least privilege
Information classificationClassification guidelines, labeling, clear desk/​screen policy, NDAs
Physical securityPerimeter security, access controls, equipment protection, secure areas, locked cabinets, electronic door locks
Vendor managementThird-party agreements, service delivery review, change management, exchange agreements, vendor register, ISO 27017 controls
Logging and monitoringAudit logs, log protection, admin logs, retention policy, Slack notifications
Development and changeSecurity requirements, secure dev guidelines, change management, separation of environments, code reviews, security tests, vulnerability scans
PersonnelManagement responsibility, awareness training, disciplinary process, offboarding, screening, confidential advisor, four-eyes principle
Media and cryptoRemovable media management, secure disposal, cryptographic controls, off-premises equipment security
ComplianceLegal compliance, IP rights, document protection, personal data protection, internal audits
Incident managementResponse procedures, reporting channels, evidence collection, lessons learned
Business continuityBC measures, cloud storage, failover, backup procedures
OperationalOwn network equipment, one-way connections, SaaS intermediary, spot checks

Agent instructions

  1. Always write in English. All field labels, values, descriptions, and measures in English.
  2. Start by understanding what the risk is about. If the user gives enough context, propose the full assessment directly. If not, ask only the questions needed.
  3. For Likelihood and Impact: if the user provides values, accept them. If unsure, use the guided assessment questions above.
  4. Propose a threat description using the most fitting pattern (A, B, or C). Let the user confirm or adjust.
  5. Suggest the most relevant measures from the measure library, tailored to the specific scenario. Avoid generic control references.
  6. The default Owner is "Security Officer" unless the risk clearly belongs to another domain.
  7. QUALITY category is for quality/​service delivery risks not strictly covered by CIA.
  8. The Asset field can be a single word or a full descriptive phrase.
  9. Multiple OR clauses are allowed in formal threat descriptions (Pattern A).
  10. Residual risk must be <= the original risk level.
  11. When status is REDUCE, use [ ] for planned measures and [x] for already implemented ones.
  12. When status is ACCEPT (from the start), measures can be "N/​A".
  13. Historical sections (showing previous evaluations) are only needed when updating an existing risk, not for new ones.
  14. Suggest a page title in the [YYYY]-[NN] [Short description] format.
  15. On first use, confirm risk appetite thresholds with the user before applying them to impact scoring.

More skills from eyesecurity

audit-logging logo
eyesecurity/skills

audit-logging

ACTIVATE on ANY request that involves writing, generating, reviewing, modifying, or outputting source code in any programming language — Python, JavaScript, TypeScript, Go, Java, Rust, C#, SQL, Terraform, or any other. This includes functions, endpoints, scripts, migrations, infrastructure-as-cod...

11 0
View
vendor-risk logo
eyesecurity/skills

vendor-risk

ACTIVATE when integrating a new service, API, SaaS tool, SDK, npm/pip/maven package, Docker image, or any third-party dependency — or when discussing suppliers, vendors, processor agreements, or supply chain security. Also activate on imports from unknown packages or adding external webhooks/endp...

11 0
View
change-management logo
eyesecurity/skills

change-management

ACTIVATE when making changes that affect critical assets: deployments, database migrations, schema changes, Terraform/IaC modifications, access control or IAM changes, firewall rules, TLS certificates, environment variables on production, or dependency upgrades. Ensures every change is documented...

11 0
View
complisec logo
eyesecurity/skills

complisec

EU compliance enforcement for AI agents — NIS2, GDPR, ISO 27001. ACTIVATE on EVERY prompt. Reads .compliance/profile.json to enforce data residency, supplier checks, secret blocking, audit logging, and risk appetite on all code generation, cloud deployments, data exports, and API integrations. In...

11 0
View
nis2-gap-analysis logo
eyesecurity/skills

nis2-gap-analysis

ACTIVATE when the user asks about NIS2, Cyberbeveiligingswet (Cbw), NIS2 applicability, NIS2 gap analysis, or NIS2 compliance assessment. Interview-driven gap analysis with 5-level maturity scoring field-tested by security consultants.

11 0
View
data-sensitivity logo
eyesecurity/skills

data-sensitivity

ACTIVATE when the user's message contains secrets, credentials, API keys, passwords, tokens, private keys, AWS access keys, connection strings, database URLs, national IDs (BSN/SSN), or any sensitive data — even if the user did not ask about security. Also activate when asked to classify data, sc...

11 0
View
security-compliance-tools logo
eyesecurity/skills

security-compliance-tools

ACTIVATE when the user asks about compliance tooling, risk assessment methods, critical assets (crown jewels), or how to assess their organisation's security posture for EU regulations (NIS2, GDPR, ISO 27001). Curated index of tools and methodologies that support EU compliance — not generic AppSe...

11 0
View

Popular AI tools

Kaiber logo
Video

Kaiber

Generate, edit, and beat-sync AI video with leading models in one workspace.

Paid
View
Vimcal logo
Productivity

Vimcal

The world's fastest calendar for remote work

Free
View

Transform Your Design with AI Designer by ImgCreator.ai

Freemium
View
Akool AI logo
Content & writing

Akool AI

Revolutionizing Video Production with AI-Powered Creativity

Paid
View

Extend an image past the frame and let AI fill the new aspect ratio.

Freemium
View
StarByFace logo
Security

StarByFace

Discover your celebrity doppelgänger with StarByFace!

Free
View
C

ChainClarity explains 700+ crypto whitepapers in plain English, with layered summaries, comparisons, research tools, alerts, and a $4.99 Pro plan.

Freemium
View
Opus Clip logo
Coding & apps

Opus Clip

Opus.ai: Revolutionize Your Web Experience

Free
View