agentleFS
peterbamuhigirepeterbamuhigire/business-plan-skills/skills/advisory-deliverables/internal-controls-and-risk-framework/SKILL.md

internal-controls-and-risk-framework

Skill4 starsChanged 18 days agoMITView on GitHub

Use when authoring an internal-control policy, risk-management framework, control matrix, or risk register for an East African organisation; use `finance-policy-and-manual` for finance procedures and defer accounting controls, audit evidence, and close doctrine to Chwezi Accounting Doctrine.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Test representative controls and exception paths.

    line 117

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Internal Controls and Risk Framework

A consulting-deliverable skill: it produces a standalone organisational document — an Internal Control & Risk Management Framework — not a business-plan section. It owns the document's structure, the consulting workflow, the control and risk parameterisation, and the East African regulatory framing — and it pulls every control design, fraud test, and risk-treatment substance from the finance engine at chwezi-accounting-doctrine.

Use When

  • A client needs an Internal Control & Risk Management Framework, Internal Control Policy, segregation-of-duties / authorisation framework, ICFR documentation pack (process narratives, risk-control matrices, walkthroughs), fraud-risk / anti-fraud policy, whistleblowing policy, or an enterprise risk register and risk-management policy.
  • Responding to an RFP/EOI for "development of an internal control and risk management framework / risk register."
  • Reviewing, gap-auditing, or maturing an existing control environment against COSO ERM / ISO 31000 and the entity's actual cycles.

Do Not Use When

  • Writing the internal-control chapter inside a finance manual — use advisory-deliverables/finance-policy-and-manual (controls are one chapter there; this skill is the full standalone framework).
  • Producing a standalone procurement manual — use advisory-deliverables/procurement-policy-and-manual.
  • Writing a business-plan risk section — use pipeline/12-risk-analysis.
  • Answering a single control-design or fraud-test question — go straight to the relevant finance-engine SKILL.md.

Required Inputs

Entity type (NGO/CSO, SME, SACCO/cooperative, donor-funded project, public/LG-adjacent body); legal form, registration (URSB/NGO Bureau), and funding model (donor-restricted, own revenue, mixed); reporting framework and whether ICFR-style assertion documentation is required; existing controls, policies, or risk register; board and committee structure (Audit/Risk committee, internal audit function); operating cycles in scope (cash, procurement, payroll, assets, grants); risk appetite signals from the board; jurisdiction (default Uganda) and whether public-sector surcharge/liability applies; named approver/owner. Never fabricate these — ask.

Workflow

  1. Intake & scope. Fix entity type, the document(s) wanted (Framework, Policy, ICFR pack, risk register), the cycles in scope, jurisdiction, and whether a public-sector surcharge overlay applies. If a framework or register exists, run a gap audit against the blueprint before drafting.
  2. Select the blueprint. Load references/document-blueprint.md for the chapter map, the control set rendered as tables, the risk-register method, and the chapter→doctrine-skill map.
  3. Pull control and risk substance from the doctrine — do not improvise. For each chapter read the mapped finance-engine SKILL.md (the blueprint carries the map), anchored on 10-controls-governance-and-fraud/internal-controls-library and sox-style-icfr-documentation, plus doctrine/references/uganda-ngo-financial-management-patterns.md (NGO) or uganda-public-sector-pfm.md (public).
  4. Draft section by section. Set every threshold, signatory count, authorisation tier, reconciliation cadence, and review frequency as a named client parameter. Propose defaults from the Uganda patterns (clearly flagged "to be board-approved"), never copied verbatim from another entity.
  5. Keep statutory rates and figures OUT of the body. Any surcharge/pecuniary-liability figures, statutory penalty rates, or tax touchpoints go into a dated, live-verified Statutory Schedule appendix that points to the source register — so the framework does not go stale.
  6. Build the control set as tables. Segregation-of-duties matrix, amount-keyed authorisation matrix, control-activity matrix by cycle (cash/bank, procurement, payroll, assets), and the bank-signatory mandate — from internal-controls-library plus the NGO/public patterns. No person occupies two adjacent links of the control chain.
  7. Build the risk-management layer. Risk register (probability × impact, owner, mitigation, monitoring), risk appetite statement, and control self-assessment — to ISO 31000 / COSO ERM, cross-referencing chwezi-sdlc-documentation 09-governance-compliance/04-risk-assessment and pipeline/12-risk-analysis for method. For public bodies, layer pecuniary-liability / surcharge per the LG (Financial & Accounting) Regulations 2007.
  8. Add fraud, ethics, monitoring, forms, and version control. Fraud-risk and forensic indicators; whistleblowing per the Whistleblowers Protection Act 2010; conflict-of-interest registers and clearance-on-exit; the internal-audit / monitoring function; the standard forms-and-registers appendix; and a review-and-amendment page with effective date and approver.
  9. Run the quality gates. meta-utility/anti-ai-slop (live), meta-strategy/meta-critical-thinking-business-logic, and the finance engine's finance-doctrine-conformance-scanner / finance-module-audit. Record each gate run in the artefact manifest.
  10. Produce the deliverable. Client-ready DOCX/PDF (use the document engines), plus an adoption checklist: board/Audit-committee resolution, effective date, control-owner assignment, and staff training plan.

Quality Bar

Every control statement is specific and testable (no "adequate controls" — name the control, owner, frequency, and evidence); segregation of duties holds even for a three-person team, with no person in two adjacent links and nobody approving in their own favour; every threshold, signatory rule, and cadence is a named parameter with an owner and review date; the risk register carries probability × impact, owner, mitigation, and monitoring against a stated appetite; statutory/surcharge figures are never hardcoded in the body; the framework maps to the entity's actual cycles and to ICFR assertions where required; the public-sector surcharge overlay is applied only where it governs; and it passes anti-ai-slop and the doctrine conformance scan. British English throughout.

Anti-Patterns

  • Vague controls ("controls are in place", "management reviews regularly") instead of named control, owner, frequency, and evidence.
  • Dropping the SoD matrix or collapsing the request→approve→disburse→account→review chain because the entity is "too small".
  • Letting one person occupy two adjacent links, approve in their own favour, or combine procurer, approver, and inventory-keeper roles.
  • Embedding surcharge, penalty, or tax figures in the body instead of the live-verified Statutory Schedule.
  • A risk register that lists risks without probability × impact, owner, mitigation, monitoring, or a stated risk appetite.
  • Copying one organisation's thresholds, signatory rules, or appetite statements verbatim into another.
  • Asserting a control or fraud-test design without reading the mapped finance-engine skill.
  • Bolting on the public-sector surcharge overlay for a private NGO, or omitting it for a body it governs.

Outputs

Internal Control & Risk Management Framework; Internal Control Policy; SoD, authorisation, and control-activity matrices; ICFR documentation pack (process narratives, risk-control matrices, walkthroughs) where required; enterprise risk register + risk-appetite statement; fraud-risk and whistleblowing policy; forms-and-registers pack (risk register, CoI register, control self-assessment checklist, exception/incident log, whistleblowing intake form); dated Statutory Schedule appendix; adoption and version-control page; quality-gate manifest.

References

  • references/document-blueprint.md — chapter map, chapter→doctrine-skill mapping, control set as tables, risk-register method, parameterisation rule, and the standard forms/registers list.
  • Finance engine (chwezi-accounting-doctrine): skills/10-controls-governance-and-fraud/ (internal-controls-library, sox-style-icfr-documentation, forensic-accounting-and-anti-fraud, whistleblowing-and-finance-ethics, aml-kyc-and-suspicious-transaction-reporting, engagement-quality-and-plain-language-output); skills/06-close-consolidation-and-reporting/audit-pbc-and-evidence-management; doctrine/references/uganda-ngo-financial-management-patterns.md, uganda-public-sector-pfm.md, uganda-compliance-caveats.md.
  • Cross-engine risk method: chwezi-sdlc-documentation/09-governance-compliance/04-risk-assessment; business-plan-skills/skills/pipeline/12-risk-analysis.
  • country-context/uganda/SKILL.md for institutions/regulatory bodies; language/east-african-english for style.
<!-- dual-compat-start -->

Inputs

ArtefactSource or providerRequired?If absent
Objectives, processes, systems, incidents, audits, and existing controlsClient records, walkthroughs, and interviewsRequiredReturn a scoped evidence request, not an invented risk assessment
Risk appetite, authority, legal, donor, and sector obligationsBoard records and verified authoritiesRequiredMark appetite and compliance scoring provisional
Finance controls and audit-evidence doctrineChwezi Accounting DoctrineRequired for money-touching risksStop finance-control conclusions pending doctrine and professional review

Outputs

ArtefactConsumerAcceptance condition
Internal Control and Risk Management FrameworkBoard, management, process owners, assurance, auditors, and fundersGovernance, risk method, control design, reporting, incidents, and review are operational
Risk register and control matrixRisk owners and assurance reviewersRisks link to causes, impacts, appetite, controls, evidence, owner, action, and residual rating

Evidence Produced

EvidenceFormatAcceptance condition
Risk-control-evidence traceabilityMatrix with test and exception fieldsEvery key control has an owner, frequency, evidence, test, failure route, and remediation date
Finance quality-gate recordDoctrine and professional-review logMoney-touching controls map to current Chwezi doctrine and unresolved review blocks certification

Capability Contract

Audit and assessment default to read-only. Edit the framework only when authorised. Do not change production controls, access rights, transactions, risk acceptance, whistleblowing records, or audit conclusions without explicit competent authority; never certify control effectiveness without test evidence.

Degraded Mode

Without walkthroughs, samples, system access, doctrine, or current legal sources, distinguish design review from operating-effectiveness testing and mark inaccessible checks not assessed. Return the narrowest risk register and evidence plan that the available facts support.

Decision Rules

ChoiceActionFailure or risk avoided
Risk exceeds appetite and no effective control existsEscalate and define a dated treatment before acceptanceUnowned material exposure
Control exists only in policyRate design only; test operation separatelyFalse assurance
Segregation is impracticalAdd documented compensating review and monitoringUncontrolled concentration of duties
Risk touches finance, tax, assets, payroll, grants, banking, or reportingApply Chwezi doctrine and professional reviewMisstated control objective or evidence

Workflow

  1. Confirm objectives, scope, users, assurance level, risk criteria, and authority.
  2. Map processes, assets, obligations, incidents, findings, systems, and dependencies.
  3. Identify risks through causes, events, impacts, existing controls, and evidence; stop where a material process cannot be observed.
  4. Score inherent risk, assess control design, and separate design from tested operation.
  5. Define treatments, owners, dates, indicators, escalation, and residual risk decisions.
  6. Apply Chwezi doctrine and qualified review to all money-touching controls and audit evidence.
  7. Test representative controls and exception paths; recover by narrowing conclusions or adding evidence work, never by upgrading an untested rating.
  8. Release with limitations, risk acceptance, action status, and review cadence visible.

Quality Standards

Risk ratings must be reasoned, controls testable, evidence retained, and limitations explicit. A control is effective only when both design and operating evidence support that conclusion; finance assertions require the Chwezi gate.

Anti-Patterns

  • Listing generic risks with no cause or impact. Fix: write a specific cause-event-impact statement.
  • Calling a policy clause an effective control. Fix: identify execution evidence and test a sample.
  • Combining ownership and independent review in one role. Fix: separate or document a compensating control.
  • Scoring every risk high to appear cautious. Fix: apply agreed criteria and record rationale.
  • Reusing a finance control without doctrine mapping. Fix: verify objective, posting, reconciliation, and evidence under Chwezi doctrine.
  • Hiding untested areas in an overall green rating. Fix: mark them not assessed and qualify the conclusion.

Worked Example

Where one staff member creates and pays suppliers, rate the segregation risk, verify whether an independent bank and supplier-master review actually operates, sample its evidence, and keep operating effectiveness unassessed if records are unavailable. For NGO cyber-resilience investment and control capacity, load NGO cyber-resilience investment conditions.

<!-- dual-compat-end -->

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.