v0.2.0 now on GitHub · EU AI Act enforcement: August 2026

Your AI agent shipped.
Did you test it?

The Preflight suite is the pre-deploy test pipeline for AI agents — regression, compliance, orchestration gates, config promotion, and observability. Five tools. One pipeline. One exit code.

npm install -g github:StanislavBG/agent-gate github:StanislavBG/agent-shift github:StanislavBG/agent-trace
View on GitHub →
~ agent-gate run && agent-shift promote staging
$ agent-gate run

    stepproof      18/18 scenarios passed      (4.2s)
    agent-comply   EU AI Act: Low risk          (1.1s)
    cost           $0.31 / run (budget: $0.50) (0.4s)

  VERDICT: PASS  — gates cleared, receipt saved

$ agent-shift promote staging --require-gate-pass ./gate-result.json

    gate receipt   verified sha256 match
    config diff    model unchanged, 1 prompt delta
    snapshot       snap-20260320-a3f1 saved

  PROMOTED  — staging → production
MIT · No telemetry · Runs offline · All gates run in parallel
The gap

Five things that silently break
every AI deploy.

01

Behavioral regression

A model update or prompt change breaks step 3 of your 7-step workflow.

You find out from a customer.
02

Compliance violation

Someone adds a model call that crosses an EU AI Act Annex III threshold.

You find out from legal.
03

Cost overrun

A prompt tweak triples token usage. No alert. No CI check.

You find out from the invoice.
04

Config drift

Staging ran with model A. Production promoted with model B. No diff, no snapshot, no rollback path.

You find out from the incident report.
05

No observability

Your agent ran. Was that 3 LLM calls or 47? Did step 5 retry? Was that 2 seconds or 20?

You find out from an unexplained invoice.

Your existing CI catches none of these. That's the gap.

The Preflight suite

Five tools, one story.

validate → comply → gate → deploy → observe. Each tool is independently useful. Together they form a complete pre-deploy pipeline for AI agents.

stepproof
Behavioral regression testing
  • YAML scenario definitions, N iterations each
  • Per-step pass rate thresholds
  • LLM-judge assertions
  • Exit 1 on regression
"Did my AI pipeline break?"
$ stepproof run ./scenarios/
GitHub →
agent-comply
EU AI Act compliance scanning
  • Scans for AI model calls across your codebase
  • EU AI Act Annex III risk classification
  • Model card & AI-SBOM generation
  • Fails on undocumented high-risk usage
"Is this legal in the EU?"
$ agent-comply report
GitHub →
Suite flagship
agent-gate
Unified pre-deploy CI gate
  • Runs stepproof + agent-comply in parallel
  • Token cost estimation & budget cap
  • Model allowlist enforcement
  • Single PASS/FAIL exit code for CI
"Is this safe to ship?"
$ agent-gate run
GitHub →
agent-shift
Config versioning & environment promotion
  • Snapshot agent config: model, prompts, tools, guardrails
  • Diff environments or snapshots side-by-side
  • Promote with optional gate-pass receipt verification
  • Rollback to any prior snapshot instantly
"Your agent passed the gates. Now promote it safely."
$ agent-shift promote staging
GitHub →
agent-trace
Local agent observability
  • OTel GenAI spans stored in local SQLite
  • Query traces from the terminal — no dashboard
  • Works offline, no cloud account required
  • SARIF export for GitHub Security tab
"What did my agent actually do?"
$ agent-trace record 'node agent.js'
GitHub →
Install the tool you need. Add the rest when you need it.
"I want to catch AI regressions"
stepproof
"I need EU compliance artifacts for legal"
agent-comply
"I want one CI gate for everything"
agent-gate
"I need safe environment promotion with rollback"
agent-shift
"I want to see what my agent did at runtime"
agent-trace
"Starting fresh — what do I install?"
agent-gate + agent-shift + agent-trace — the full pipeline
How it works

One CI step. Five tools. Ship with confidence.

validate → comply → gate → deploy → observe. The first three checks run in parallel. agent-shift closes the loop. agent-trace records what happened.

git push
Triggers CI
stepproof
~4.2s
18 scenarios, pass rate thresholds, regression exit
agent-comply
~1.1s
Annex III scan, model card, AI-SBOM
$
cost check
~0.4s
Token estimate vs. budget, model allowlist
agent-gate
PASS/FAIL
Unified verdict + gate receipt JSON
agent-shift
~0.2s
Snapshot config, verify receipt, promote
deploy
Confident push with full audit trail
stepproof, agent-comply, and cost check run in parallel via Promise.allSettled. agent-shift only runs after agent-gate exits 0 — no gate receipt, no promotion.
~ gate failure + config drift
$ agent-gate run

    stepproof      14/18 scenarios failed
     → stepproof run --verbose to debug step failures

    agent-comply   EU AI Act: Low risk          (1.1s)

    cost           $1.82 / run (budget: $0.50) — OVER BUDGET
     → Check for prompt expansion in recent commits

  VERDICT: FAIL  — 2 of 3 gates blocked deploy

$ agent-shift check

    config drift   staging ≠ production
     → model: gpt-4o-mini (staging) vs gpt-4o (production)
     → Run: agent-shift diff staging production
     → Run: agent-shift rollback production   (to revert)
Drop this into your CI. Done.
# .github/workflows/preflight.yml
name: Preflight Agent Check
on: [push, pull_request]

jobs:
  preflight:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: '20'

      - name: Install Preflight suite
        run: |
          npm install -g github:StanislavBG/stepproof
          npm install -g github:StanislavBG/agent-comply
          npm install -g github:StanislavBG/agent-gate
          npm install -g github:StanislavBG/agent-shift
          npm install -g github:StanislavBG/agent-trace

      - name: Run Preflight gates
        run: agent-gate run --config .preflight.json
        # Exits 1 on regression, compliance failure, or budget breach

      - name: Promote config to staging
        if: github.ref == 'refs/heads/main'
        run: agent-shift promote staging --require-gate-pass ./gate-result.json
        # No gate receipt → no promotion. Enforced.

Why five tools, not one?

Each tool solves a distinct, independently valuable problem. stepproof tests behavior. agent-comply scans for legal risk. agent-gate unifies both for CI. agent-shift handles what comes after the gate: promoting the validated config to production with a rollback path if something goes wrong post-deploy. agent-trace is the observability layer: wrap any agent run to record OTel spans in local SQLite — no cloud account, no dashboard, data stays on your machine.

Passing a gate is a point-in-time verdict. Promoting a config is a stateful operation with environment history, diffs, and rollback. Observing production behavior closes the loop: agent-trace traces wire back to the stepproof scenarios that defined expected behavior. You need all five — and they wire together. agent-gate writes a receipt; agent-shift verifies it; agent-trace records what actually ran.

Pricing

Free at the CLI. Paid at the team layer.

All four CLI tools are MIT, forever. The cloud dashboard — for team compliance history, PDF artifacts, and audit exports — is where we charge.

Free
$0 always
OSS CLI — all five tools
  • stepproof + agent-comply + agent-gate + agent-shift + agent-trace
  • Local execution, zero telemetry
  • MIT License
  • Full scan, test, gate, snapshot, promote
  • JSON + Markdown artifact output
  • Community YAML policy library
  • CI/CD integration (GitHub Actions, etc.)
$ npm i -g github:StanislavBG/agent-gate github:StanislavBG/agent-shift github:StanislavBG/agent-trace
Enterprise
$299/mo or custom
For teams shipping AI to regulated markets
  • Everything in Pro
  • SSO / SAML
  • Full promotion + rollback audit trail for EU regulators
  • Custom policy authoring + regulatory review
  • SLA + priority support
  • Unlimited seats · private cloud option
Contact us →
"What you hand your DPO when the regulator asks for AI governance evidence — gate verdicts, config snapshots, and full promotion history."
CLI execution is free — forever. The moment you gate the CLI, a competitor forks it.

Cloud dashboard coming.

Team compliance history, shared audit trails, PDF reports for legal, agent-shift promotion timelines. Join the list to get early access.

No spam. One email when cloud launches.
Get started

Run your first check
in 60 seconds.

Install any tool and run it against your codebase. No account. No config file required to start.

npm install -g github:StanislavBG/agent-gate
then: agent-gate run