coso-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@coso-mcp create a COSO engagement and compute sample size for quarterly controls"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
coso-mcp
An open Model Context Protocol server for internal-control reviews under the COSO Internal Control—Integrated Framework (ICIF-2013). It covers the full review lifecycle — scope → risk assessment → control register → population → sampling → testing → workpapers → review sign-off → deficiency evaluation → report — grounded in the framework's 5 components, 17 principles and points of focus.
Local: runs over stdio (
npx @simonmak-ascent/coso-mcp).Hosted: public endpoint at https://internal-control.ascent-partners.com/mcp (reference + read tools, plus an ephemeral per-session sandbox for the full write lifecycle — nothing persists).
Not audit or legal advice. This server assists with internal-control review methodology. All workpapers and assessments must be reviewed by qualified professionals before use in audit documentation.
Why this exists (competitive advantage)
Existing MCP servers either expose a vendor's GRC records (read-mostly,
proprietary, product-shaped) or cover ISO 27001 / security controls. None runs
a traceable COSO review. coso-mcp is the first COSO/ICFR-native,
write-capable, open MCP for the whole lifecycle.
Dimension | coso-mcp | Vendor GRC MCPs (Workiva, Vanta, Drata, Optro, OpenPages, LogicGate…) | Community MCPs (iso27001-mcp, unidoc/isms, mcp-server-scf…) | AI skills (sox-testing / audit-support) |
COSO ICIF-2013 native (5 components / 17 principles / points of focus) | ✅ core engine | ❌ product objects | ❌ ISO/SCF-shaped | ⚠️ text only |
Full lifecycle to signed workpaper | ✅ | ❌ retrieval; Workiva GRC tools are read-only | ❌ planning/findings at best | ⚠️ templates, not executable |
Write/stateful workpapers + preparer/reviewer sign-off + audit trail | ✅ | ❌ mostly read | ❌ | ❌ |
Framework effectiveness (principle/component present & functioning) | ✅ | ❌ | ❌ | ❌ |
Deterministic, reproducible computation (sample-size tables, seeded selection) | ✅ | partial | ❌ | ❌ (freeform) |
Open source + open access (MIT, public endpoint, no account) | ✅ | ❌ paid platform | ⚠️ some OSS | ⚠️ not a server |
Moat: it is the lifecycle layer, not the data layer — it can sit on top of Workiva/Vanta/Drata rather than compete for their data; its sampling and deficiency math are computed, not improvised; and openness gives zero-friction distribution no gated vendor can match.
Related MCP server: contractOracle
Tools
Tool | Kind | Purpose |
| read | Server identity, framework coverage, disclaimer. |
| read | The five COSO components. |
| read | The 17 principles (optionally filtered by component). |
| read | A principle by number or slug, with paraphrased points of focus. |
| read | A component by id, with its principles. |
| read | Operations / reporting / compliance. |
| read | The CEAVOP assertions used to frame control objectives. |
| write | Create a review engagement. |
| read | Get an engagement by id. |
| read | List engagements. |
| write | Add a significant account/process to scope. |
| read | List scope items. |
| write | Register a risk to objectives (likelihood, assertions, principles, fraud). |
| read | List registered risks. |
| write | Register a control (type, frequency, linked risks, assertions). |
| read | List registered controls. |
| read | Append-only audit events for an engagement. |
| read | Recommended sample size from frequency × risk, with rationale. |
| write | Register a population (item list or declared size, bounded). |
| read | List populations. |
| write | Seeded, reproducible sample selection (random/systematic/targeted/haphazard). |
| write | Record attribute-level outcomes (pass/fail/na) and a conclusion. |
| read | List test results. |
| write | Record an exception with root cause and impact. |
| read | List exceptions. |
| write | Assemble a Markdown control-testing workpaper. |
| read | List workpapers. |
| read | Get a workpaper with its Markdown. |
| write | Add a reviewer note. |
| write | Resolve a reviewer note. |
| write | Preparer/reviewer sign-off (warns on same-person independence). |
| read | Append-only activity log for an engagement. |
| write | Classify a deficiency (deficiency / significant / material weakness). |
| read | List deficiencies. |
| read | Group and flag deficiencies that may aggregate. |
| read | Coverage indicators for one COSO principle. |
| read | Coverage indicators for one COSO component. |
| read | Check all five components have coverage. |
| read | Generate the Markdown review report. |
Resources: coso://framework/icif-2013 (framework reference),
coso://engagement/{id} (engagement bundle: scope, risks, controls).
Prompts: scope-review, assess-risk, build-control-matrix,
draft-test-workpaper, evaluate-deficiencies.
Usage
Local (stdio)
npx -y @simonmak-ascent/coso-mcpClaude Desktop / generic MCP client config:
{
"mcpServers": {
"coso": {
"command": "npx",
"args": ["-y", "@simonmak-ascent/coso-mcp"]
}
}
}Hosted (streamable HTTP)
https://internal-control.ascent-partners.com/mcpThe hosted endpoint is open. Write operations run in an ephemeral, isolated, rate-limited per-session sandbox and are not persisted.
Development
Compute runs on a box, never locally: cs run "<cmd>".
pnpm install
pnpm typecheck
pnpm test
pnpm test:coverage
pnpm build # tscSecurity
See SECURITY.md. Report vulnerabilities privately — do not open a public issue.
Contributing
See CONTRIBUTING.md and the Code of Conduct.
License
MIT © Simon Mak / Ascent Partners Group Limited.
Available Tools
39 toolsadd_review_noteAdd a review noteCInspect
Add a reviewer note to a workpaper.
Inputs: text (required), workpaperId (required), engagementId (required), author. Returns: note, workpaper
add_scope_itemAdd a scope itemCInspect
Add a significant account/process to the review scope, with its objective category and assertions.
Inputs: area (required), engagementId (required), objectiveCategory (required), inScope, +2 more. Returns: scopeItem
aggregate_deficienciesAggregate deficienciesCInspect
Group related deficiencies by control, flag combinations that may aggregate, and report the overall severity.
Inputs: engagementId (required), deficiencyIds. Returns: aggregation
assemble_workpaperAssemble a workpaperCInspect
Assemble a Markdown control-testing workpaper from the control, its risks, population, sample, test results and exceptions.
Inputs: controlId (required), engagementId (required), title, testResultId. Returns: workpaper
assess_componentAssess a COSO component (indicators)BInspect
Report coverage indicators for one COSO component, aggregating its principles.
Inputs: id (required), engagementId (required). Returns: found, assessment
assess_principleAssess a COSO principle (indicators)CInspect
Report data-driven coverage indicators for one COSO principle (controls, risks and deficiencies mapped to it).
Inputs: number (required), engagementId (required). Returns: found, assessment
classify_deficiencyClassify a deficiencyBInspect
Classify a control deficiency as deficiency / significant deficiency / material weakness from likelihood × magnitude (escalates for a pervasive control).
Inputs: title (required), magnitude (required), likelihood (required), description (required), +7 more. Returns: deficiency
compute_sample_sizeCompute a sample sizeBInspect
Recommend a sample size for a control from its frequency, risk and population size, with a documented rationale.
Inputs: frequency (required), risk, populationSize, priorDeficiency, +1 more. Returns: capped, frequency, rationale, recommended, +3 more
create_engagementCreate an engagementAInspect
Create a new internal-control review engagement. Returns the engagement id used by all later tools.
Inputs: title (required), entity (required), id, period, +1 more. Returns: engagement
create_populationCreate a populationBInspect
Register the population for a control — either an explicit item list or a declared size. Bounded to 10000 items.
Inputs: description (required), engagementId (required), items, period, +3 more. Returns: population
framework_integration_checkFramework integration checkBInspect
Check that all five COSO components have some recorded coverage and that principles are addressed.
Inputs: engagementId (required). Returns: components, integration
generate_reportGenerate the review reportBInspect
Generate a Markdown internal-control review report: scope, risks, controls, testing, workpapers, deficiencies and COSO coverage.
Inputs: engagementId (required). Returns: counts, markdown
get_activity_logGet the activity logBInspect
Return the append-only activity log for an engagement, plus a Markdown rendering.
Inputs: engagementId (required). Returns: count, events, markdown
get_audit_logGet the audit logCInspect
Return the append-only audit events recorded for an engagement.
Inputs: engagementId (required). Returns: count, events
get_componentGet a COSO componentAInspect
Return a single COSO component by id, with its principles.
Inputs: id (required). Returns: found, component
get_engagementGet an engagementCInspect
Return an engagement by id.
Inputs: id (required). Returns: found, engagement
get_principleGet a COSO principleCInspect
Return a single COSO principle by framework number or slug, including its paraphrased points of focus.
Inputs: id, number. Returns: found, principle
get_server_infoAInspect
Return the server identity, the COSO framework it covers, and the mandatory disclaimer.
Inputs: no named parameters. Returns: name, title, version, framework, +2 more
get_workpaperGet a workpaperCInspect
Return a workpaper including its rendered Markdown.
Inputs: workpaperId (required), engagementId (required). Returns: found, workpaper
list_assertionsList audit assertions (CEAVOP)AInspect
List the assertions used to frame control objectives: completeness, existence/occurrence, accuracy, valuation, obligations/rights, presentation/disclosure.
Inputs: no named parameters. Returns: count, assertions
list_componentsList COSO componentsAInspect
List the five components of the COSO Internal Control—Integrated Framework (ICIF-2013).
Inputs: no named parameters. Returns: count, framework, components
list_controlsList controlsCInspect
List the controls registered for an engagement.
Inputs: engagementId (required). Returns: count, controls
list_deficienciesList deficienciesCInspect
List the deficiencies recorded for an engagement.
Inputs: engagementId (required). Returns: count, deficiencies
list_engagementsList engagementsAInspect
List all engagement ids and titles.
Inputs: no named parameters. Returns: count, engagements
list_exceptionsList exceptionsCInspect
List the exceptions recorded for an engagement.
Inputs: engagementId (required). Returns: count, exceptions
list_objective_categoriesList COSO objective categoriesAInspect
List the three COSO objective categories (operations, reporting, compliance).
Inputs: no named parameters. Returns: count, objectiveCategories
list_populationsList populationsCInspect
List the populations registered for an engagement.
Inputs: engagementId (required). Returns: count, populations
list_principlesList COSO principlesAInspect
List the 17 principles of the COSO ICIF-2013, optionally filtered to a single component.
Inputs: component. Returns: count, framework, principles
list_risksList risksCInspect
List the risks registered for an engagement.
Inputs: engagementId (required). Returns: count, risks
list_scope_itemsList scope itemsCInspect
List the scope items for an engagement.
Inputs: engagementId (required). Returns: count, scopeItems
list_test_resultsList test resultsCInspect
List the test results recorded for an engagement.
Inputs: engagementId (required). Returns: count, testResults
list_workpapersList workpapersCInspect
List the workpapers assembled for an engagement.
Inputs: engagementId (required). Returns: count, workpapers
note_exceptionNote an exceptionBInspect
Record an exception identified during testing, with root cause, impact and any compensating control.
Inputs: reference (required), description (required), engagementId (required), testResultId (required), +3 more. Returns: exception
record_test_resultRecord a test resultCInspect
Record attribute-level outcomes (pass/fail/na) for a tested sample, with an optional conclusion.
Inputs: results (required), sampleId (required), attribute (required), controlId (required), +5 more. Returns: testResult
register_controlRegister a controlCInspect
Register a key control with its type, frequency, linked risks, assertions and COSO principles.
Inputs: type (required), frequency (required), description (required), engagementId (required), +6 more. Returns: control
register_riskRegister a riskCInspect
Register a risk to objectives, with likelihood, assertions, related COSO principles and a fraud flag.
Inputs: description (required), engagementId (required), cause, fraud, +5 more. Returns: risk
resolve_review_noteResolve a review noteCInspect
Mark a review note resolved on a workpaper.
Inputs: noteId (required), workpaperId (required), engagementId (required), resolvedBy. Returns: note, workpaper
select_sampleSelect a sampleBInspect
Select a reproducible sample from a population. Seeded (deterministic) for random/systematic/haphazard; targeted takes explicit references. Records the method, seed and references.
Inputs: method (required), engagementId (required), populationId (required), seed, +4 more. Returns: sample
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 39 tools
Most tools map to a clearly distinct resource+action in the audit lifecycle (engagement, scope, risk, control, population, sample, test result, exception, workpaper, deficiency, report). The main overlap is get_audit_log vs get_activity_log, both described as append-only per-engagement logs, and the classify/aggregate/list_deficiencies trio requires reading descriptions to separate. Overall boundaries are clear enough that an agent can select correctly.
Nearly all tools follow a consistent snake_case verb_noun pattern (create_engagement, register_risk, select_sample, list_exceptions). Deviations are minor: sign_off is verb-only, framework_integration_check puts the verb last, and assessment tools use assess_/classify_/aggregate_ prefixes that still read consistently.
39 tools is heavy for a single MCP server, and the surface is inflated by many near-duplicate list_* tools (list_risks, list_controls, list_scope_items, list_populations, list_test_results, list_exceptions, list_workpapers, list_deficiencies) that could be one parameterized list endpoint. Each tool is individually defensible for a full audit lifecycle, but the set is larger than the workflow strictly requires.
The surface covers the whole engagement lifecycle end-to-end: setup, scoping, risk/control registration, sampling, testing, exceptions, workpapers, review notes, sign-off, deficiency classification, COSO coverage assessment, and reporting. Gaps are minor: there are no update/delete operations for engagements, risks, controls, or scope items, so corrections must be appended rather than amended.
Maintenance
Related MCP Connectors
Compliance frameworks (SOC 2, ISO 27001, CMMC, NIST, more) delivered to AI agents as MCP tools.
Your audit methodology inside Claude: findings, risks, controls and workpapers in your team format.
10,065 source-verified compliance nodes, 39 pillars, 25 MCP tools (EU AI Act, GDPR, NIST, MITRE).
One MCP server for all your MCP servers: code mode, per-tool permissions, approvals and audit.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables DORA Article 8 change management through MCP tools for registering and tracking changes, assessing risk, validating approvals, rollback plans, and test evidence, checking patch compliance and segregation of duties, and viewing change statistics and server health.-
- FlicenseNot gradedqualityDmaintenanceEnables DORA-compliant ICT contract management through 10 MCP tools, including mandatory clause checking, critical/important function analysis, subcontracting chain assessment, exit readiness evaluation, and risk scoring.-
- FlicenseNot gradedqualityDmaintenanceEnables DORA ICT governance and compliance workflows through 10 MCP tools for managing audit findings, generating board reports, reviewing frameworks, tracking controls, exceptions, remediation actions, KPIs, and annual compliance evidence.-
- AlicenseAqualityBmaintenanceEnables AI agents to screen journal entries against 11 rules, prioritize them by risk score, and run balance checks, duplicate detection, and Benford analysis via MCP.573 npm3MIT