MeshFlow
  • Cloud PlatformDeployIntegrationsInit CLIWhy MeshFlow
  • DocsLearn
  • CustomersCertificationBlogChangelogCommunity
  • Pricing
Sign inGet started
Pricing
Sign inGet started
MeshFlow

The production-safe standard for agentic AI.
Apache 2.0 — free forever.

Product
  • Pricing
  • Cloud
  • Compare
  • Init CLI
  • Changelog
Docs
  • Quickstart
  • Core concepts
  • Compliance guide
  • Token optimization
  • API reference
Resources
  • Blog
  • Customers
  • Community
  • PyPI
  • GitHub
Company
  • Discord
  • Twitter
  • Security
  • License
MeshFlow © 2026 · Yaya Systems · Apache 2.0
All systems operational
MeshFlowdocs
Search docs...⌘K
v1.13.0 GitHub Discord
Getting startedQuickstartInstallationCore conceptsHarness architectureSkills vs ToolsSandbox mode
GuidesBuilding agentsMulti-agent workflowsModel RoutersCost AnalyticsCompliance guideToken optimization
Claude & AIAdvisor patternThinking budgetsDynamic workflowsContext compactionTool streaming
Governancemeshflow-forensicZero-Trust PoliciesSOC 2 assertionEU AI ActProduction Readiness
IntegrationsOpenAI Agents SDKLangGraphCrewAIMCP ToolsTypeScript SDKOTel / OTLPAutoGen 0.4+
API ReferenceWorkflowAgentCostCapAdvisorAgentCostRegressionGate
Docs/Governance/Production Readiness

Production Readiness Certification

meshflow certify scores your agents across 8 governance categories and issues a cryptographically signed report. Use it in CI to gate deployments on a minimum readiness score.

What it checks
Every check pulls live evidence from your MeshFlow workspace — trace spans, checkpoints, context packs, compliance reports, eval runs, and worker jobs. No self-reporting.

Install

bash
pip install "meshflow[certify]"
# or
npm install @meshflow/certify

Run

Run certification against your org. Optionally scope to a single run:

python
import meshflow

report = meshflow.certify(org_id="my-org")
print(f"Score: {report.overall_score:.0%} Verdict: {report.verdict}")

# Scope to a single run
report = meshflow.certify(org_id="my-org", run_id="run_abc123")

Or use the CLI directly:

bash
meshflow certify --org my-org
meshflow certify --org my-org --run run_abc123

CI gate

Fail the build when agents fall below a minimum score. Pass --min-score to set the threshold:

yaml
# .github/workflows/certify.yml
- name: Certify agents
run: |
pip install "meshflow[certify]"
meshflow certify --org $MESHFLOW_ORG --min-score 0.85
env:
MESHFLOW_API_KEY: $${ secrets.MESHFLOW_API_KEY }
Non-zero exit on failure
meshflow certify exits 1 when the score is below --min-score or the verdict is not_ready. This causes the CI step to fail automatically.

Scoring rubric

Overall score is a weighted average. Verdicts: Production Ready ≥ 85%, Needs Work 65–84%, Not Ready < 65%.

CategoryWeightWhat it checks
Traceability Ledger20%Spans emitted, input/output hashes, causal chain, model version
Policy Compliance20%Zero violations, policy assigned to all runs, compliance reports
Replayability15%Checkpoints written, failed runs replayed, step-indexed state
Context Provenance15%Context packs emitted, pack hashes, source inventory
Tool Safety10%Tool versions tracked, guardrails active
Eval Coverage10%Eval suite exists, pass rate ≥ 80%
Cost Governance5%Cost tracked per run, no budget overruns
Incident Recovery5%No unresolved paused runs, worker queue healthy

Evidence bundle

Every report can be exported as a signed JSON evidence bundle — useful for SOC 2, EU AI Act, or internal review boards. The bundle includes all span, run, and checkpoint records plus a SHA-256 bundleHash for tamper detection.

bash
# Download from the dashboard
GET /api/dashboard/certify/{id}/export

# CLI shorthand
meshflow certify export --report-id rpt_abc123 --out ./meshflow-cert.json

The bundle is self-contained: share it with auditors without granting MeshFlow access.

Previous
OTel / OTLP
On this pageInstallRunCI gateScoring rubricEvidence bundle