PromptSpeak Governance
This adds a validation layer between Claude and any MCP tools it tries to call. It provides 56 governance tools that intercept tool calls before execution, running them through a 9-stage pipeline that checks structural validity, security patterns, and behavioral drift. You get a hold queue for human approval of risky operations, security scanning that catches SQL injection and hardcoded secrets before file writes, and an audit trail of every attempted action. It uses PromptSpeak frames (symbolic sequences like ⊕◊▶α) to encode governance constraints. Reach for this when you need deterministic guardrails around agentic workflows, especially in regulated domains where you can't afford to debug production incidents caused by hallucinated arguments or prompt injection attacks.
- tools
- 43
- transport
- STDIO
- last updated
- Mar 17, 2026
Tools
43 toolsPublic tool metadata for what this MCP can expose to an agent.
ps_validateValidate a frame against all rules without executing
Validate a frame against all rules without executing
No parameter schema in public metadata yet.
ps_validate_batchBatch validation for efficiency
Batch validation for efficiency
No parameter schema in public metadata yet.
ps_executeFull pipeline: validate → hold check → execute → audit
Full pipeline: validate → hold check → execute → audit
No parameter schema in public metadata yet.
ps_execute_dry_runRun full pipeline without executing the tool
Run full pipeline without executing the tool
No parameter schema in public metadata yet.
ps_hold_listList all operations awaiting human approval
List all operations awaiting human approval
No parameter schema in public metadata yet.
ps_hold_approveApprove with optional modified arguments
Approve with optional modified arguments
No parameter schema in public metadata yet.
ps_hold_rejectReject with reason
Reject with reason
No parameter schema in public metadata yet.
ps_hold_configConfigure hold triggers and thresholds
Configure hold triggers and thresholds
No parameter schema in public metadata yet.
ps_hold_statsHold queue statistics
Hold queue statistics
No parameter schema in public metadata yet.
ps_state_getGet agent's active frame and last action
Get agent's active frame and last action
No parameter schema in public metadata yet.
ps_state_systemSystem-wide statistics
System-wide statistics
No parameter schema in public metadata yet.
ps_state_haltTrip circuit breaker — blocks all future calls
Trip circuit breaker — blocks all future calls
No parameter schema in public metadata yet.
ps_state_resumeReset circuit breaker
Reset circuit breaker
No parameter schema in public metadata yet.
ps_state_resetFull state reset
Full state reset
No parameter schema in public metadata yet.
ps_state_drift_historyDrift detection alert history
Drift detection alert history
No parameter schema in public metadata yet.
ps_delegateCreate parent→child delegation with constrained permissions
Create parent→child delegation with constrained permissions
No parameter schema in public metadata yet.
ps_delegate_revokeRemove delegation
Remove delegation
No parameter schema in public metadata yet.
ps_delegate_listList active delegations
List active delegations
No parameter schema in public metadata yet.
ps_config_setSet configuration key-value pairs
Set configuration key-value pairs
No parameter schema in public metadata yet.
ps_config_getGet current config
Get current config
No parameter schema in public metadata yet.
ps_config_activateActivate a named configuration
Activate a named configuration
No parameter schema in public metadata yet.
ps_config_exportExport full config as JSON
Export full config as JSON
No parameter schema in public metadata yet.
ps_config_importImport config from JSON
Import config from JSON
No parameter schema in public metadata yet.
ps_confidence_setSet confidence thresholds
Set confidence thresholds
No parameter schema in public metadata yet.
ps_confidence_getGet current thresholds
Get current thresholds
No parameter schema in public metadata yet.
ps_confidence_bulk_setBatch threshold update
Batch threshold update
No parameter schema in public metadata yet.
ps_feature_setEnable/disable specific checks
Enable/disable specific checks
No parameter schema in public metadata yet.
ps_feature_getGet feature flags
Get feature flags
No parameter schema in public metadata yet.
ps_symbol_createCreate symbol with type, metadata, and tags
Create symbol with type, metadata, and tags
No parameter schema in public metadata yet.
ps_symbol_getRetrieve by ID
Retrieve by ID
No parameter schema in public metadata yet.
ps_symbol_updateUpdate metadata or tags
Update metadata or tags
No parameter schema in public metadata yet.
ps_symbol_listList with optional type filter
List with optional type filter
No parameter schema in public metadata yet.
ps_symbol_deleteDelete by ID
Delete by ID
No parameter schema in public metadata yet.
ps_symbol_importBatch import
Batch import
No parameter schema in public metadata yet.
ps_symbol_statsRegistry statistics
Registry statistics
No parameter schema in public metadata yet.
ps_symbol_formatFormat symbol for display
Format symbol for display
No parameter schema in public metadata yet.
ps_symbol_verifyMark symbol as verified
Mark symbol as verified
No parameter schema in public metadata yet.
ps_symbol_list_unverifiedList symbols needing verification
List symbols needing verification
No parameter schema in public metadata yet.
ps_symbol_add_alternativeAdd alternative identifier
Add alternative identifier
No parameter schema in public metadata yet.
ps_security_scanScan content, return findings by severity
Scan content, return findings by severity
No parameter schema in public metadata yet.
ps_security_gateScan + enforce: block/hold/warn/allow
Scan + enforce: block/hold/warn/allow
No parameter schema in public metadata yet.
ps_security_configList, enable, disable, change severity of patterns
List, enable, disable, change severity of patterns
No parameter schema in public metadata yet.
ps_audit_getFull audit trail with filters
Full audit trail with filters
No parameter schema in public metadata yet.
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageReadme
promptspeak-mcp-server
Pre-execution governance for AI agents. Blocks dangerous tool calls before they execute.
AI agents call tools (file writes, API requests, shell commands) with no validation layer between intent and execution. A prompt injection, hallucinated argument, or drifting goal can trigger irreversible actions. PromptSpeak intercepts every MCP tool call, validates it against deterministic rules, and blocks or holds risky operations for human approval — in 0.1ms, before anything executes.
PromptSpeak Governance Demo
When to use this
Details
- Repository
- github.com/chrbailey/promptspeak-mcp-server
- Package
@chrbailey/promptspeak-mcp-server- Transport
- STDIO
- Registry
- MCP registry, active