Skip to content

MCP_TOOLS.md

Latest commit

 

History

History
1033 lines (748 loc) · 48.2 KB

MCP_TOOLS.md

File metadata and controls

1033 lines (748 loc) · 48.2 KB
Raw

SyncroNow AI MCP Tool Reference

This reference is generated from packages/mcp-server/src/toolSchemas.ts. Do not edit it manually; regenerate with node packages/mcp-server/scripts/generate-tool-reference.js.

Total tools: 63.

Contents

sync_ tools

sync_status

Show connected ServiceNow instance, scope and user from current SyncroNow AI project.

  • Version: 1.0.0
Parameter Type Required Default Description
logLevel string (one of: "error", "warn", "info", "debug", "silly") no "info"
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Captured result of the underlying syncrona status CLI run. Only success results (exit code 0) carry structuredContent.

Field Type Always present Description
exitCode integer yes Process exit code of the CLI run (always 0 on success results).
timedOut boolean yes True when the CLI run was killed after exceeding timeoutMs.
stdout string yes Captured standard output of the CLI run.
stderr string yes Captured standard error of the CLI run.
correlationId string no Request correlation id injected by the server.

sync_get_session_context

Get current ServiceNow session context: active scope and active update set for current user.

  • Version: 1.0.0
Parameter Type Required Default Description
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Current ServiceNow session context for the connected user.

Field Type Always present Description
userSysId string yes sys_id of the connected sys_user record.
scope object yes Active application scope.
updateSet object yes Active update set from the sys_update_set user preference.
correlationId string no Request correlation id injected by the server.

sync_set_scope

Set current ServiceNow scope for the current user session using scope code (for example x_nuvo_sinc).

  • Version: 1.2.0
  • Safety: supports dryRun
Parameter Type Required Default Description
scope string yes
dryRun boolean no false Plan the scope switch and record an audit entry without changing the session.
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Scope-switch confirmation with the refreshed session context. When dryRun is requested, the payload is the dry-run plan (dryRun, tool, planned) instead.

Field Type Always present Description
requestedScope string no Scope code that was requested.
scopeSysId string no sys_id of the resolved sys_scope record.
scopeName string no Display label of the resolved scope.
sessionContext object no Refreshed session context (same shape as the sync_get_session_context output).
dryRun boolean no Present and true only on dry-run results.
tool string no Tool that produced the plan (dry-run results only).
planned object no Arguments the tool would apply (dry-run results only).
correlationId string no Request correlation id injected by the server.

sync_list_scopes

List available scopes from sys_scope. Use optional encoded query to filter.

  • Version: 1.0.0
Parameter Type Required Default Description
query string no ""
limit number (min 1, max 500) no 100
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Page of sys_scope rows matching the query.

Field Type Always present Description
count integer yes Number of rows returned in this page.
truncated boolean no True when more rows match than the limit returned.
rows array<object> yes sys_scope rows limited to the sys_id, scope and name fields.
correlationId string no Request correlation id injected by the server.

sync_set_update_set

Set current ServiceNow update set for current user by name or sys_id. Optionally create missing set by name.

  • Version: 1.2.0
  • Safety: supports dryRun
Parameter Type Required Default Description
updateSetName string no ""
updateSetSysId string no ""
createIfMissing boolean no true
dryRun boolean no false Plan the update-set switch and record an audit entry without changing the session.
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Update-set switch confirmation with the refreshed session context. When dryRun is requested, the payload is the dry-run plan (dryRun, tool, planned) instead.

Field Type Always present Description
targetUpdateSet object no Update set that is now active for the current user.
sessionContext object no Refreshed session context (same shape as the sync_get_session_context output).
dryRun boolean no Present and true only on dry-run results.
tool string no Tool that produced the plan (dry-run results only).
planned object no Arguments the tool would apply (dry-run results only).
correlationId string no Request correlation id injected by the server.

sync_list_update_sets

List update sets from sys_update_set. Use optional encoded query to filter.

  • Version: 1.0.0
Parameter Type Required Default Description
query string no ""
limit number (min 1, max 500) no 100
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Page of sys_update_set rows matching the query.

Field Type Always present Description
count integer yes Number of rows returned in this page.
truncated boolean no True when more rows match than the limit returned.
rows array<object> yes sys_update_set rows limited to the sys_id, name, state, application and sys_created_on fields.
correlationId string no Request correlation id injected by the server.

sync_prepare_session

Ensure expected scope and update set are active. It reads current context, applies changes when needed, and returns final context.

  • Version: 1.2.0
  • Safety: supports dryRun
Parameter Type Required Default Description
expectedScope string no ""
expectedUpdateSetName string no ""
expectedUpdateSetSysId string no ""
createUpdateSetIfMissing boolean no true
dryRun boolean no false Plan the session alignment and record an audit entry without changing the session.
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Report of the session changes that were applied, with before and after context. When dryRun is requested, the payload is the dry-run plan (dryRun, tool, planned) instead.

Field Type Always present Description
actions array<string> no Human-readable descriptions of the changes that were applied.
changed boolean no True when at least one change was applied.
initialContext object no Session context before changes (same shape as the sync_get_session_context output).
finalContext object no Session context after changes (same shape as the sync_get_session_context output).
dryRun boolean no Present and true only on dry-run results.
tool string no Tool that produced the plan (dry-run results only).
planned object no Arguments the tool would apply (dry-run results only).
correlationId string no Request correlation id injected by the server.

sync_preflight_check

Validate current scope/update-set against guardrail expectations in sync.mcp.guardrails.json or provided overrides.

  • Version: 1.0.0
Parameter Type Required Default Description
expectedScope string no ""
expectedUpdateSetName string no ""
expectedUpdateSetSysId string no ""
timeoutMs number (min 1000, max 900000) no

sync_check_instance_capabilities

Check whether SyncroNow AI scoped-app endpoints are available in the target ServiceNow instance.

  • Version: 1.0.0
Parameter Type Required Default Description
scope string no "" Optional scope code for manifest check. Defaults to current session scope.
timeoutMs number (min 1000, max 900000) no

sync_refresh

Refresh local SyncroNow AI manifest from the target instance.

  • Version: 1.0.0
Parameter Type Required Default Description
logLevel string (one of: "error", "warn", "info", "debug", "silly") no "info"
timeoutMs number (min 1000, max 900000) no

Output (structuredContent on success results):

Captured result of the underlying syncrona refresh CLI run. Only success results (exit code 0) carry structuredContent.

Field Type Always present Description
exitCode integer yes Process exit code of the CLI run (always 0 on success results).
timedOut boolean yes True when the CLI run was killed after exceeding timeoutMs.
stdout string yes Captured standard output of the CLI run.
stderr string yes Captured standard error of the CLI run.
correlationId string no Request correlation id injected by the server.

sync_build

Build local files with SyncroNow AI.

  • Version: 1.0.0
Parameter Type Required Default Description
diff string no "" Optional branch name for --diff build
logLevel string (one of: "error", "warn", "info", "debug", "silly") no "info"
timeoutMs number (min 1000, max 900000) no

sync_push

Push files to ServiceNow instance using SyncroNow AI. Destructive action.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
target string no "" Optional encoded target path(s); must not start with "-"
diff string no "" Optional branch name for --diff
scopeSwap boolean no false
updateSet string no ""
logLevel string (one of: "error", "warn", "info", "debug", "silly") no "info"
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean yes false Must be true to execute push.
dryRun boolean no false

sync_create_script_include

Create sys_script_include record, then optionally run sync refresh so the file is downloaded locally.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
name string yes
apiName string no ""
script string no ""
active boolean no true
clientCallable boolean no false
refreshAfterCreate boolean no true
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean yes false
dryRun boolean no false

sync_create_script_include_and_sync

Create Script Include, refresh manifest/files, and return local candidate file paths for immediate editing.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
name string yes
apiName string no ""
script string no ""
active boolean no true
clientCallable boolean no false
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean yes false
dryRun boolean no false

sync_detect_drift

Compare local and instance record states and report drift summary with actions.

  • Version: 1.0.0
Parameter Type Required Default Description
localRecords array<object> no []
instanceRecords array<object> no []
updateSetSysId string no ""

sync_validate_change_package

Validate that selected records include required dependencies.

  • Version: 1.0.0
Parameter Type Required Default Description
selectedIds array<string> yes []
graph object yes

sync_build_semantic_index

Build semantic symbol index from local source files.

  • Version: 1.0.0

This tool has no input parameters.

sync_search_semantic_index

Search current semantic symbol index.

  • Version: 1.0.0
Parameter Type Required Default Description
query string no ""

sync_symbol_cross_reference

Summarize semantic symbol cross-references by symbol name, files, and occurrences.

  • Version: 1.0.0

This tool has no input parameters.

sync_health_check

Return MCP health, endpoint diagnostics timeline, and per-tool reliability metrics.

  • Version: 1.0.0
Parameter Type Required Default Description
timeoutMs number (min 1000, max 900000) no

sync_metrics_trend

Compute trend deltas between previous and current diagnostics windows.

  • Version: 1.0.0

This tool has no input parameters.

sync_tool_contract_info

Return tool-contract version, declared tool list, and deterministic contract hash for capability negotiation.

  • Version: 1.0.0

This tool has no input parameters.

sync_table_api_coverage_matrix

Return current metadata/object coverage matrix and supported operations via Table API.

  • Version: 1.0.0

This tool has no input parameters.

sync_plan_minimal_footprint

Rank where-to-modify targets for a task using graph context and minimal-footprint scoring.

  • Version: 1.0.0
Parameter Type Required Default Description
task string yes
graph object yes
limit number (min 1, max 20) no 5

sync_ai_next_actions

Generate prioritized AI next actions and recommended tool calls from a natural-language objective.

  • Version: 1.0.0
Parameter Type Required Default Description
objective string yes
maxSteps number (min 1, max 10) no 5
currentContext object no {}
constraints object no {}

sync_generate_scope_knowledge

Generate scope knowledge artifacts (md + json) and optionally write them under .syncrona-mcp/scopes/.

  • Version: 1.0.0
  • Safety: supports dryRun
Parameter Type Required Default Description
scope string no ""
entities array<object> no []
graph object no {}
risks array<object> no []
suppressions array<object> no []
updateSetContext object no {}
task string no ""
writeFiles boolean no false
trigger string (one of: "manual", "init", "refresh", "successful_change", "drift") no "manual"
dryRun boolean no false Plan the run and record an audit entry without writing any file to the workspace.

sync_generate_scope_docs

Generate full scope documentation bundle under .syncrona-mcp/docs/{scope}/ with overview, dependencies, table relationships, and per-object docs.

  • Version: 1.0.0
  • Safety: supports dryRun
Parameter Type Required Default Description
scope string no ""
entities array<object> no []
graph object no {}
task string no "scope docs"
includeFields boolean no true
includeDiagrams boolean no true
includeScheduledJobs boolean no true
includeCrossScope boolean no true
writeFiles boolean no false
dryRun boolean no false Plan the run and record an audit entry without writing any file to the workspace.

sync_validate_scope_knowledge

Validate scope knowledge json payload against required schema fields.

  • Version: 1.0.0
Parameter Type Required Default Description
index object yes

sync_scope_knowledge_auto_update

Run trigger-based scope knowledge update contract for init/refresh/successful_change/drift.

  • Version: 1.0.0
  • Safety: supports dryRun
Parameter Type Required Default Description
trigger string (one of: "init", "refresh", "successful_change", "drift") yes
scope string no ""
entities array<object> no []
graph object no {}
task string no ""
writeFiles boolean no false
dryRun boolean no false Plan the run and record an audit entry without writing any file to the workspace.

sync_generate_table_dependency_report

One-command table dependency report generation with deterministic output paths under .syncrona-mcp/reports/.

  • Version: 1.0.0
  • Safety: supports dryRun
Parameter Type Required Default Description
scope string no ""
task string no "table dependencies report"
writeFiles boolean no false
dryRun boolean no false Plan the run and record an audit entry without writing any file to the workspace.

sync_analyze_scope_relations

Analyze scope tables and relations (explicit, hidden, inferred) from ServiceNow metadata and local workspace evidence.

  • Version: 1.0.0
Parameter Type Required Default Description
scope string no ""
includeWorkspace boolean no true
includeServiceNow boolean no true
timeoutMs number (min 1000, max 900000) no

sync_onboarding_bootstrap

Return onboarding wizard/checklist with sensible defaults and quickstart guidance.

  • Version: 1.0.0

This tool has no input parameters.

sync_unified_change_workflow

Run one-command workflow shell: preflight, deep analysis, minimal-footprint check, approval gate, and optional apply.

  • Version: 1.2.0
  • Safety: mutating - gated by confirmDestructive; supports dryRun
Parameter Type Required Default Description
task string yes
script string no ""
taskType string (one of: "script", "metadata", "hybrid") no "hybrid"
executionMode string (one of: "mocked", "remote") no "mocked"
allowRemoteApply boolean no false
remoteScript string no ""
remoteEndpoint string no ""
proposedChanges array<object> no []
footprintBudget object no {}
riskLevel string (one of: "low", "medium", "high", "critical") no
approval object no {}
rollbackEvidence object no {}
policy object no {}
apply boolean no false
confirmDestructive boolean no false
writeSimulationReport boolean no false
simulationId string no ""
dryRun boolean no false Plan the workflow and record an audit entry without applying anything. Takes precedence over apply: a dry run never mutates the instance.
timeoutMs number (min 1000, max 900000) no

sync_list_recent_changes

List recent changes in a scope from sys_update_xml since a timestamp (default 24h), grouped by record.

  • Version: 1.0.0
Parameter Type Required Default Description
scope string yes Scope code to filter changes (for example x_nuvo_sinc).
since string no ISO timestamp lower bound. Defaults to 24 hours ago.
limit number (min 1, max 200) no 50
timeoutMs number (min 1000, max 900000) no

sync_generate_release_notes

Generate release notes from an Update Set's sys_update_xml records, grouped by table, in markdown or json.

  • Version: 1.0.0
Parameter Type Required Default Description
updateSetSysId string no sys_id of the update set. Provide this or updateSetName.
updateSetName string no Name of the update set. Used when updateSetSysId is not supplied.
format string (one of: "markdown", "json") no "markdown"
timeoutMs number (min 1000, max 900000) no

sync_run_atf_tests

Trigger ATF test execution in the instance (a single test, a suite, or all suites in a scope) and poll for pass/fail results per step.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
scope string yes Scope code that owns the ATF tests (for example x_nuvo_sinc).
suiteId string no sys_id of an ATF test suite to run.
testId string no sys_id of a single ATF test to run. Cannot be combined with suiteId or runAll.
runAll boolean no false Run all ATF suites in the scope. Used when suiteId and testId are omitted.
confirmDestructive boolean yes false Required acknowledgement — running ATF tests executes a background script and mutates the instance. Must be true to run.
dryRun boolean no false Plan the run and record an audit entry without triggering ATF execution.
timeoutMs number (min 1000, max 900000) no

sync_cicd_run

Run an ATF suite or test, or install, publish or roll back an application, through the ServiceNow CI/CD REST API (sn_cicd) — the same requests as syncrona cicd <action>. Dispatches the action, polls its progress tracker until it finishes or timeoutMs runs out, and for ATF actions reads the linked result. Returns outcome succeeded | failed | incomplete with the CLI's exitCode 0 | 2 | 1, plus the verdict and reason of syncrona cicd --json: verdict passed (exit 0); failed or no_tests (exit 2; no_tests is a suite that ran nothing, which is not a pass); incomplete (exit 1, the run could not be followed to its end) or unknown (exit 1, the tracker finished but its ATF result is unreadable or unclear). reason says why, in full and possibly over several lines (fenced as untrusted instance data): the ATF summary or the tracker's end state on a pass, the instance's reason or what failed otherwise, and the message when the run is incomplete or unknown. An incomplete run returns its progressId; pass it back as progressId to resume polling that tracker without dispatching new work. A rejected call (an unknown action, an invalid progressId or parameter, or a dispatch without confirmDestructive=true) dispatches nothing and returns only an error text, with no outcome, verdict, reason or exitCode. Needs the sn_cicd.sys_ci_automation role.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
action string (one of: "run-suite", "run-test", "install", "publish", "rollback") yes The CI/CD action to run.
suiteId string no run-suite: sys_id of the ATF test suite (exactly one of suiteId or suiteName).
suiteName string no run-suite: name of the ATF test suite (exactly one of suiteId or suiteName).
testId string no run-test: sys_id of the ATF test (required).
browserName string no run-suite: browser to schedule the run on (for example chrome).
browserVersion string no run-suite: browser version.
osName string no run-suite: operating system name.
osVersion string no run-suite: operating system version.
runInCloud boolean no run-suite / run-test: run on the Cloud Runner.
performance boolean no run-suite: run as a performance run.
captureNodeLogs boolean no run-test: capture node logs for the run.
scope string no install / publish / rollback: application scope (exactly one of scope or appSysId).
appSysId string no install / publish / rollback: application sys_id (exactly one of scope or appSysId).
appVersion string no install / publish: version to install or publish. rollback: the version to roll back to (required).
baseAppVersion string no install: base application version.
autoUpgradeBaseApp boolean no install: upgrade the base application automatically.
devNotes string no publish: developer notes for the published version.
progressId string no Resume instead of dispatch: the 32-character hexadecimal sys_id of an existing sn_cicd progress tracker (the progressId an incomplete run returned). Polls that tracker — and for ATF actions reads its result — with the same outcome mapping; no dispatch request is sent, so confirmDestructive may be false, and the dispatch arguments are ignored. action is still required: it picks the ATF result record to read.
pollMs number (min 250, max 60000) no 1000 Interval between progress-tracker polls, in milliseconds.
confirmDestructive boolean yes false Required acknowledgement — every action dispatches work that changes the instance (application installs, publishes and rollbacks, or ATF runs that write results). Must be true to dispatch; a progressId resume only polls and accepts false.
dryRun boolean no false Return the request that would be dispatched (or, with progressId, the tracker that would be polled) and record an audit entry without calling the instance.
timeoutMs number (min 1000, max 900000) no Budget for the whole call (dispatch, polling and result read). When it runs out the outcome is incomplete and the work may still be running on the instance.

sync_fluent_build

Build a ServiceNow Fluent project (a directory holding now.config.json) through the optional @syncrona/fluent package and the ServiceNow SDK — the same build as syncrona fluent build. It never calls the instance, but it is a mutating tool (the tool policy and the mutating audit apply; as a local write it is exempt from the instance preflight that enforcePreflightForMutations turns on, unless the tool's own policy sets requirePreflight): it overwrites the project's dist/ output, and it executes the project's installed @syncrona/fluent and @servicenow/sdk and any build-time code the project pulls in inside the server process, with the server's full environment (including instance credentials) and no sandbox — run it only on a project you trust as much as running its build yourself. An adapter resolved from the project whose real path leaves the workspace (a symlinked package or node_modules, or one hoisted above the workspace) is never loaded from there: the server's own install (its node_modules ancestry, never NODE_PATH or a global folder) is used when it has the adapter, and otherwise the call is refused with code FLUENT_ADAPTER_OUTSIDE_WORKSPACE. The response and the audit name the adapter that was loaded: adapter.source (project | server), adapter.path (the real path of its package directory) and adapter.version. Returns the build errors and warnings, the output paths, and outcome succeeded | failed | incomplete with the CLI's exitCode 0 | 2 | 1. When @syncrona/fluent or @servicenow/sdk is not installed, returns outcome incomplete with code FLUENT_NOT_INSTALLED or FLUENT_SDK_MISSING and an install hint. Errors and warnings are capped at 200 each (errorCount and warningCount give the totals, errorsTruncated and warningsTruncated mark a cut); symlinks in the output are not followed, and an output directory that cannot be listed is reported in outputWarnings without failing the build.

  • Version: 1.0.0
  • Safety: supports dryRun
Parameter Type Required Default Description
project string no "." Fluent project directory, relative to the workspace; it must contain now.config.json and may not lead outside the workspace (symlinks included).
frozenKeys boolean no false Fail when generated keys would change (the CLI's --frozen-keys).
errorOnConflict boolean no false Treat key conflicts as errors (the CLI's --error-on-conflict).
skipClean boolean no false Keep the previous output directory instead of cleaning it (the CLI's --skip-clean).
dryRun boolean no false Plan the run and record an audit entry without loading the adapter, running any project code or writing any file to the workspace.
timeoutMs number (min 1000, max 900000) no Budget for the build — best-effort, not a hard limit. The build runs in-process and cannot be cancelled: when the budget runs out the outcome is incomplete and the build may still write its output. A build that blocks the event loop (synchronous compile work) delays the timeout, and the server, until it returns; such a run returns its real result with budgetExceeded: true.

sync_validate_before_push

Pre-push validation pipeline: runs security/architecture analysis on a scope's scripts, checks for recent conflicting changes, and reports ready or blocked per record.

  • Version: 1.0.0
Parameter Type Required Default Description
scope string yes Scope code to validate (for example x_nuvo_sinc).
tables array<string> no Optional subset of script tables to validate.
limit number (min 1, max 200) no 50
conflictWindowHours number (min 1, max 720) no 24
timeoutMs number (min 1000, max 900000) no

sync_compare_instances

Compare a scope's script records between two stored instance profiles (for example dev vs prod) and report records that are only in one side or differ by content.

  • Version: 1.0.0
Parameter Type Required Default Description
profileA string yes First stored instance profile name (as saved by syncrona login).
profileB string yes Second stored instance profile name.
scope string yes Scope code to compare.
tables array<string> no Optional subset of script tables to compare.
limit number (min 1, max 500) no 200
timeoutMs number (min 1000, max 900000) no

sync_export_update_set

Export an Update Set as XML via the export_update_set processor and return the XML plus metadata; optionally writes it under .syncrona-mcp/exports.

  • Version: 1.0.0
Parameter Type Required Default Description
updateSetSysId string no sys_id of the update set. Provide this or updateSetName.
updateSetName string no Name of the update set. Used when updateSetSysId is not supplied.
writeFiles boolean no false When true, write the XML to .syncrona-mcp/exports/{name}.xml.
timeoutMs number (min 1000, max 900000) no

sync_suggest_tests

Generate an ATF (Automated Test Framework) server-side test skeleton from a Script Include. Fetches the Script Include script from the instance (unless script is provided inline), analyzes its public methods, and returns a ready-to-paste ATF test script plus import instructions.

  • Version: 1.0.0
Parameter Type Required Default Description
scope string no Scope code of the Script Include (for example x_nuvo_sinc). Defaults to current session scope.
scriptIncludeName string yes Name of the Script Include to generate tests for.
scriptIncludeSysId string no Optional sys_id of the Script Include. Used instead of scriptIncludeName when provided.
script string no Optional Script Include source. When provided, the instance is not queried and this source is analyzed directly.
timeoutMs number (min 1000, max 900000) no

sync_diff_instance_vs_local

Compare local scoped files against the records on the instance. Fetches records from the given table/scope, compares them with the local copies, and reports changed, added (local-only) and removed (instance-only) records with a diff summary and race-condition warnings.

  • Version: 1.0.0
Parameter Type Required Default Description
scope string no Scope code to compare (for example x_nuvo_sinc). Defaults to current session scope.
tableName string no Table to compare. Defaults to sys_script_include.
recordName string no Optional single record name to limit the comparison to.
timeoutMs number (min 1000, max 900000) no

run_ tools

run_workspace_command

Run a local command in the SyncroNow AI workspace for automation tasks.

  • Version: 1.2.0
  • Safety: mutating - gated by confirmDestructive; supports dryRun
Parameter Type Required Default Description
command string yes Executable name (for example: node, npm, npx)
args array<string> no []
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean no false Set true if running potentially destructive commands such as deploy/download.
dryRun boolean no false Plan the invocation and record an audit entry without running it.

run_node_code

Execute JavaScript code with Node.js in the project workspace.

  • Version: 1.2.0
  • Safety: mutating - gated by confirmDestructive; supports dryRun
Parameter Type Required Default Description
code string yes
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean no false Set true to acknowledge that executing Node.js code may modify workspace or environment state.
dryRun boolean no false Plan the invocation and record an audit entry without running it.

sn_ tools

sn_query_records

Query records from a ServiceNow table using sysparm_query and optionally analyze grouped counts. Results beyond the limit cap are not lost: fetch further pages with the offset parameter.

  • Version: 1.0.0
Parameter Type Required Default Description
table string yes
query string no ""
fields array<string> no []
limit number (min 1, max 500) no 50
offset integer (min 0) no 0 Number of rows to skip before returning results (maps to sysparm_offset). Combine with limit to paginate through tables larger than the 500-row limit cap. Rows are ordered by sys_id after any ORDERBY in query, so consecutive pages neither repeat nor skip records.
analyzeField string no "" Optional field name for grouped count analysis
timeoutMs number (min 1000, max 900000) no

sn_create_record

Create a record in an allowlisted ServiceNow table. By default only scoped-app artifact tables (the metadata registry tables plus sys_script_include) are allowed; extra tables can be permitted via the SYNCRONA_MCP_CREATE_TABLE_ALLOWLIST environment variable (comma-separated), while high-risk system tables such as sys_user, sys_user_has_role, sys_properties and cmdb_ci stay denied regardless.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
table string yes
record object yes
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean yes false
dryRun boolean no false

sn_execute_background_script

Execute ServiceNow background script and return raw output/result for analysis.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
script string yes
endpointPath string no "" Optional API path for custom script execution endpoint. Default: /api/x_nuvo_sinc/sinc/runBackgroundScript
timeoutMs number (min 1000, max 900000) no
confirmDestructive boolean yes false
dryRun boolean no false

sn_list_metadata_records

List key ServiceNow metadata records with normalized schema (BR, Client Script, ACL, Dictionary, UI Policy, Scripted REST).

  • Version: 1.0.0
Parameter Type Required Default Description
recordType string (one of: "business_rule", "client_script", "ui_script", "ui_action", "ui_formatter", "acl", "dictionary", "ui_policy", "scripted_rest", "scheduled_job") yes
query string no ""
limit number (min 1, max 500) no 100
timeoutMs number (min 1000, max 900000) no

sn_get_metadata_record

Get one metadata record by sys_id and normalize output.

  • Version: 1.0.0
Parameter Type Required Default Description
recordType string (one of: "business_rule", "client_script", "ui_script", "ui_action", "ui_formatter", "acl", "dictionary", "ui_policy", "scripted_rest", "scheduled_job") yes
sysId string yes
timeoutMs number (min 1000, max 900000) no

sn_update_metadata_record

Controlled update for core metadata records with dry-run and confirmation gate.

  • Version: 1.0.0
  • Safety: mutating - requires confirmDestructive: true; supports dryRun
Parameter Type Required Default Description
recordType string (one of: "business_rule", "client_script", "ui_script", "ui_action", "ui_formatter", "acl", "dictionary", "ui_policy", "scripted_rest", "scheduled_job") yes
sysId string yes
updates object yes
confirmDestructive boolean yes false
dryRun boolean no false
timeoutMs number (min 1000, max 900000) no

sn_build_dependency_graph

Build dependency graph from script/metadata records and inferred references.

  • Version: 1.0.0
Parameter Type Required Default Description
records array<object> no []

sn_analyze_impact

Return ranked impact if target graph node changes.

  • Version: 1.0.0
Parameter Type Required Default Description
graph object yes
targetId string yes

sn_diff_dependency_graphs

Compare before/after dependency graphs and return deterministic added/removed nodes and edges.

  • Version: 1.0.0
Parameter Type Required Default Description
beforeGraph object yes
afterGraph object yes

sn_analyze_script_architecture

Run architecture anti-pattern analysis for a script.

  • Version: 1.0.0
Parameter Type Required Default Description
script string yes

sn_analyze_script_security

Run security-focused static analysis for a script.

  • Version: 1.0.0
Parameter Type Required Default Description
script string yes

sn_analyze_script_performance

Run performance-focused static analysis for a script.

  • Version: 1.0.0
Parameter Type Required Default Description
script string yes

sn_analyze_script_full

Unified architecture/security/performance script analysis with suppression and weighted risk summary.

  • Version: 1.0.0
Parameter Type Required Default Description
script string yes
suppressedIds array<string> no []
policy object no {}
nowIso string no ""

sn_autonomous_remediation_workflow

Detect, propose patch, dry-run/apply, and validate script remediations with approval gate.

  • Version: 1.0.0
  • Safety: mutating - gated by confirmDestructive; supports dryRun
Parameter Type Required Default Description
script string yes
apply boolean no false
dryRun boolean no true
confirmDestructive boolean no false

sn_render_analysis_markdown

Render unified full analysis report as deterministic markdown output.

  • Version: 1.0.0
Parameter Type Required Default Description
report object yes

sn_search_scripts

Full-text search across ServiceNow script tables (script includes, business rules, client scripts, UI scripts, scripted REST, transform scripts) and return matches with excerpts. The result carries searchComplete; if any table query fails the call returns an error result, so matchCount: 0 only means "no matches" when searchComplete is true.

  • Version: 1.1.0
Parameter Type Required Default Description
query string yes Text to search for inside script fields.
scope string no Optional scope code to restrict the search.
tables array<string> no Optional subset of script tables to search. Defaults to all of them. Any name outside the supported set is reported back in unknownTables and forces searchComplete: false — it is never silently ignored.
limit number (min 1, max 100) no 20
timeoutMs number (min 1000, max 900000) no

sn_get_record_history

Get the change history (sys_audit) for a single record, including field-level old/new values.

  • Version: 1.0.0
Parameter Type Required Default Description
table string yes Table name of the audited record.
sysId string yes sys_id of the record to inspect.
limit number (min 1, max 200) no 20
timeoutMs number (min 1000, max 900000) no

jira_ tools

jira_get_issue

Fetch rich context for the Jira issue you are working on (summary, description, status, type, priority, assignee/reporter, labels, components, parent, subtasks, linked issues, fix versions, and recent comments). Supports Jira Cloud and Server/Data Center. Provide an issue key, or omit it to infer the key from the current git branch name.

  • Version: 1.0.0
Parameter Type Required Default Description
key string no Jira issue key (for example PROJ-123). If omitted, the key is inferred from the current git branch name.
profile string no Jira credential profile to use (default: default).
comments number (min 0) no Number of most-recent comments to include (default 5; 0 to omit).
logLevel string (one of: "error", "warn", "info", "debug", "silly") no "info"
timeoutMs number (min 1000, max 900000) no