Skip to main content
Research buildEmployer demand remains unvalidated · this preview tests the product direction

Intra-org Skill attribution Product preview

Make the people behind your AI leverage visible.

Register employee-authored AI Skills. Verify real reuse. Close a transparent reward program without changing how teams work.

  • No wallet
  • No raw prompts
  • No data saved

Illustrative program

Northstar Systems

Draft close

Employer-funded pool

$5,000

July 2026Review required
Registered Skills
14
Active Creators
8
Outcome-backed uses
37
Recent evidenceSimulated
0142Pull Request Risk BriefPayments · accepted resultIncluded
0141Incident HandoffSupport · linked incidentIncluded
0140Support Escalation TriageSuccess · outcome pendingPending
Evidence, not pay-per-callInspect protocol proof →

The operating loop

From authored work to a defensible reward close.

The chain is optional plumbing. The product is the monthly system of record your organization can explain.

  1. 01Registry

    Know what exists.

    Import an approved Skill with its Creator, maintainer, version, and lineage intact.

    Pull Request Risk Brief · v1.4.0
  2. 02Evidence

    Prove real reuse.

    Record successful use inside existing tools, then attach acceptance and downstream outcomes.

    3 Wielders · 2 teams · 1 linked PR
  3. 03Reward close

    Reward contribution.

    Apply a reviewable policy to a fixed employer-funded pool and export a plain-language statement.

    Provisional · reviewed · payroll-ready

One ledger · three jobs

Useful to the whole buying group.

The Creator is the center of the promise. The employer funds the program. The platform team makes the evidence trustworthy.

For AI platform teams

A governed portfolio, not a folder of prompts.

See adoption, versions, owners, failures, and duplicate work across the organization.

For Creators

Durable credit for work that keeps working.

See who reuses your Skill, where it creates value, and how the reward policy treats that evidence.

For Total Rewards

A defensible monthly close.

Set budgets and policy once, review exceptions, resolve disputes, and export approved awards.

No-wallet product sandbox

See the entire attribution loop before connecting anything.

Illustrative sampleNo account, network call, payment, or saved data.
  1. 01RegisterCreator + version
  2. 02Record use3 teammates
  3. 03Add outcomeAccepted + linked
  4. 04Preview closeFixed pool

Sandbox ready. Import the fictional sample Skill to begin.

Skill record · sampleReady to import

Approved internal Skill

Pull Request Risk Brief

Turns a pull-request diff into a rollout and risk brief.

Creator
Maya ChenAI Enablement
Version
v1.4.09c4a9d4e…f121
Source
northstar/ai-skills/skills/pull-request-risk-brief/SKILL.md
Beneficiary
Northstar SystemsFictional organization
Version evidence stays attached to this artifact hash. The sandbox does not contact the sample repository.
Evidence ledger · local preview0/3 uses

No teammate evidence yet.

Register the Skill, then add three deterministic sample uses. Nothing leaves this browser tab.

Next action

Creates a local demo record. No repository is contacted.

Proof stays attached

The product direction changed. The receipts did not.

The repository keeps measured, modeled, and hypothetical claims separate. The employer workflow above is the next hypothesis to test.

64settlement invariants

deterministic fork and reward arithmetic

97clone-harness checks

offline, reproducible, and dependency-free

48settled x402 calls

real testnet payment-overhead sample

Open the manifesto + x402 proof Inspect the source

Testnet evidence · not a traction claim

Founding design partners · 3 openings

Run one real Skill portfolio through one real reward close.

  • 8–12 week paid pilot
  • One Skill source and one runtime
  • Creator + admin evidence views
  • Policy preview and award-ready export
Discuss a pilot