Minimal surface
Your MCP client receives one callable schema, not the full HumanMirror catalogue.
humanmirror_do →ONE TOOL · MCP · VERIFIED OUTCOMES
One tool. Less context. Verified results.
HumanMirror One gives an AI agent a single MCP tool — humanmirror_do. It can quote a measurable outcome for free, execute it through HumanMirror Outcome, independently verify the result and keep the charge only when success is proven.
Dry run
Set dry_run: true to get the deterministic verifier, contract ID and price before execution.
Verified success
One executes the Outcome contract. Failed execution or failed verification is automatically refunded.
Execute →One schema. Ten verified result types.
clean_jsonStable, idempotently cleaned JSON.
deduplicate_recordsZero remaining duplicates.
normalize_entityStable canonical entity.
data_quality_at_leastMeasured quality threshold.
detect_anomaliesDeterministic anomaly result.
no_secretsZero credential-like findings.
output_qualitySentinel pass and trust threshold.
schema_compatibleControlled breaking-change count.
safe_preflightOperational risk below threshold.
execution_receiptTrace hash proof matches payload.
The entire MCP configuration.
Remote MCP:
https://humanmirror.fr/api/one/mcp/
Tool:
humanmirror_do
Authorization:
Bearer hm_nexus_...HumanMirror One is intentionally narrow: it does not pretend subjective or high-impact objectives are deterministically provable. Unsupported objectives are rejected and not charged.