Public proof surface

Evidence reconciled through 20 July 2026

Release scope: Architecture, Achievements, Knowledge Sharing

Source: GPT KB + Git

Curated static release — not a continuous live-status feed

Release: AIOS profile v0.2 + Governance layer update

Back to org roles
Capability detailExecution acceleration layer

Runner / Super Runner

Runner / Super Runner executes scoped work inside approved boundaries. It coordinates handoff, evidence capture, and escalation. It does not approve, verify itself, or expand scope.

Role type

Execution and coordination skill

Primary shift

Ad hoc task execution to bounded evidence-aware execution

Core output

Changed artifact or handoff packet with receipts

Boundary

Not Gate, final reviewer, approval authority, or self-verifier

Why this role exists

AI-assisted execution can become fast but unsafe if scope, handoff, evidence, and claim boundaries are not explicit. Runner exists to make execution faster while preserving accountability, traceability, and escalation discipline.

What was upskilled

Before

  • ad hoc task execution
  • unclear handoff boundaries
  • speed without consistent receipt structure
  • worker routing handled manually case-by-case

After

  • scoped execution discipline
  • evidence-before-claim behavior
  • reproducible handoff packet creation
  • stop-state detection
  • escalation routing
  • same-page role-detail explanation pattern
  • readable evidence and non-claim presentation
Permanent skill pack
runner-scoped-execution

Executes only the approved slice and avoids scope expansion.

runner-handoff-and-evidence-capture

Creates artifacts, receipts, validation notes, and missing-evidence disclosure.

super-runner-coordination-skill

Coordinates multi-step execution inside boundaries without becoming authority.

Operating principles
Scope before speedEvidence before claimRunner is not GateSmall bounded slice firstReproducible handoffStop when state is unclearUse the right worker for the workSpeed must not erase accountability
External references and how they are used
DORA / delivery performance

Used as a learning foundation for speed with stability, recovery, and delivery discipline.

Google SRE

Used as a learning foundation for stop-state, reliability, incident/recovery thinking, and operational discipline.

OpenTelemetry-style thinking

Used as an adapted reference model for trace, log, evidence, correlation, and receipt mindset.

Diataxis

Used as a learning foundation for handoff-readable documentation: how-to, reference, explanation, and tutorial-style guidance.

OWASP / NIST-style risk awareness

Used as an adapted learning foundation for boundary, security, data exposure, and claim-risk awareness.

How this role applies the skill

Source-of-truth lock - scope boundary - task packet - worker routing - execution - validation evidence - missing-evidence disclosure - escalation or gate handoff - receipt.

Inputs / Outputs

Inputs

  • owner intent
  • source of truth
  • allowed scope
  • forbidden additions
  • worker/task packet
  • expected artifact
  • validation requirement

Outputs

  • changed artifact or handoff packet
  • receipt
  • route ledger
  • telemetry/validation note
  • missing evidence list
  • non-claims preserved
  • next safe action
Escalates to

Gate / Owner

approval, claim, release, or risk decisions are required

Checker / QA

deterministic validation and evidence review are required

Big Crew

architecture, implementation split, maintainability, handover, scalability, or cross-functional requirements appear

Surface Runner / Surface Guild

rendered proof, public-surface readability, or UI evidence is required

Evidence

Available Evidence

  • Runner / Super Runner local dedicated detail route exists.
  • Provider-backed review exists only for the review/spec layer if applicable.
  • Local validation receipts exist for the bounded implementation layer.

Missing Evidence

  • No final A+ grade unless separately evidenced.
  • No rendered proof unless accepted visual proof exists.
  • No live verification unless deployed and verified.
  • No production readiness claim.

Claims Not Made

  • Runner is not Gate.
  • Runner is not final reviewer.
  • Runner is not approval authority.
  • Runner does not verify itself.
  • No A+ claim by default.
  • No live verification claim.
  • No production readiness or framework certification claim.
Upskill Path / Promotion Evidence
LevelLabelEvidence to promote
1Scoped Runnercan execute one bounded task without scope expansion
2Evidence Runnercan produce receipt, validation summary, and missing-evidence list
3Super Runner Coordination Skillcan coordinate multi-step execution while preserving authority boundary
4Execution System Designercan turn Runner pattern into reusable role-detail template and handoff discipline