NAIF GRAVITY / DEVELOPER LAB

SMALL FIX. VISIBLE PROOF.

A fix you can
actually verify.

Run the failure. Inspect the mapping. Replay the same inputs and see what changed.

No sign-inRuns in your browserDownload the evidence

A synthetic webhook example. Your real issue gets its own scope and eligibility check.

mapping.specBROWSER ONLY
Synthetic input
{
  "body": {
    "order_id": "DEMO_A",
    "quantity": 2
  }
}
Current expression
$json.order_id
FIXTURE RUNNER2 synthetic cases
○event_A—
○event_B—
READY

Same inputs. Before and after. Start with the failing test.

Built around the failure you can reproduce.
WEBHOOKSMCPGITHUB ACTIONS
For developers · Start with the failure

Get the smallest fix that unblocks your build.

Choose a reproducible problem. Review the inputs, deliverables and current price before requesting paid work.

Identify the problem, try the free diagnostic, then review the repair scope and price.

NAIF Patch Evidence — Verify delivered source files · $9 one-time فحص مطابقة الملفات المستلمة محليًا · 9 دولارات دفعة واحدة.

01 / WEBHOOK

Event arrived. Fields are empty.

For an n8n field that moved under body or data.

Bring
Current paths + 2 synthetic examples.
Receive
Corrected mapping, n8n expressions and before/after checks.
Build a free test case →Scope & current priceAutomated supported mapping repair. You apply the result.
02 / MCP & BACKEND

One reproducible code failure.

For a server or request-handling defect with a clear expected result.

Bring
Minimal code + sanitized error + expected behavior.
Receive
A scoped patch, regression check and delivery notes.
Locate the failure first →Repair scope & current priceHuman-assisted service. Scope review required.
03 / GITHUB ACTIONS

Build or test keeps failing.

For one failing workflow with enough logs to reproduce the problem.

Bring
Workflow YAML + failing job logs.
Receive
A bounded workflow patch and validation notes.
See a worked example →Try the free actionlint checkerHuman-assisted service. No repository access required to browse.
Only need a basic check? Start with free tools.

NAIF free diagnostics · jq for JSON validation · MCP Inspector

Paid help makes sense when you need a scoped repair and its evidence. Syntax checks alone may be enough for a simple issue.

Evidence-first · public beta

Bring your own reproducible failure.

Start with free triage. Review the matching service, its exact scope and current price before paid work.

COMING SOON · NAIF AGENT EVOLUTION

Make every failure teach your next test — before any real action.

Meet Mission Arena ($9 / 30 days) and Future Shield ($29 / 30 days). Planned Replay Duel comparisons, Bug Vaccine synthetic regression tests and Ghost Gate preflight checks for agents and small teams.

Private Shadow prototypes only. No subscription purchase, external agent execution, wallet enrollment or live protection is available yet.

Explore the two upcoming plans
Public services · transparent pricing

Choose the problem area.

Buy each service with a one-time payment at the displayed price. Open the service catalog to compare scope and delivery time. Enterprise Assurance remains separate and unpriced.

View services & prices

ENTERPRISE

Review code changes
before merging.
راجع تغييرات الشيفرة
قبل دمجها.

Three connected products for teams evaluating AI-generated code changes. Start with a private, scoped pilot built around evidence your team can review.

01 / ANALYZEV2.5.7

NAIF RDR

Understand what changed.

Evaluate supported code changes for behavioral differences and give reviewers concrete evidence within an agreed scope.

  • Scoped behavioral analysis
  • Counterexample when found
  • Evidence for technical review
The analysis foundation
02 / CONTROLv0.4 · Enterprise Canary

NAIF Agent Mutation Firewall

Put a decision before the merge.

Apply a review gate to supported agent-proposed changes, with a clear outcome and evidence attached to the decision.

  • ALLOW / QUARANTINE / UNSUPPORTED
  • Unsupported stays unresolved
  • Mutation Passport and receipt
The mutation review gate
03 / EVALUATEv1

NAIF AMF Pilot Kit

See it in your PR workflow.

Evaluate the stack on agreed GitHub pull requests in a separate sandbox, with checks and a reviewable pilot report.

  • GitHub PR checks and evidence
  • Read-only / check-only evaluation
  • Artifacts, receipts and pilot report
Enterprise Pilot Ready
A private pilot. A defined scope. Reviewable evidence.

For companies and technical teams. Supported Python changes only; scope and acceptance criteria are agreed before evaluation.

Enterprise Access — Coming Soon

Institutional availability by announcement.

For AI agents

Machine-readable from discovery to delivery.

Agents can discover capabilities, run a bounded free probe, request a quote and retrieve supported results with a per-job token.

GET /.well-known/agent-card.json (A2A)
GET /api/agent/catalog
GET /.well-known/naif-agent.json
GET /api/commerce/policy
POST /api/agent/probe
POST /api/agent/quote
For companies

Private evidence, not marketing claims.

Enterprise buyers can use private proof rooms, configurable retention, explicit heavy-work handoff and evidence-linked delivery passports.

Enterprise Access — Coming Soon
Institutional availability by announcement.

Trust

What we claim is limited to what we can evidence.

Published delivery evidence

0 delivered jobs · 0 delivery passports. These public counters are loaded from the trust endpoint; the interactive examples above are separate synthetic tests.

No SOC 2 or ISO 27001 certification is claimed. Secrets are not exposed by the storefront. Heavy external work is never started automatically.

Constrained execution

Server executors run with CPU, memory, timeout and filesystem limits.

Delivery passport

Delivered jobs can carry QA and artifact-hash evidence in one record.

Private proof rooms

Enterprise quotes can be shared through token-protected rooms without exposing job artifacts.