ao2-rsi-operatorlisted

# AO2 RSI Operator Use this skill to run and interpret AO2 recursive self-improvement evidence. It does not publish, approve, or claim `full_autonomous_self_mutating_rsi`. ## Claim Boundary - Supported claim: `bounded_governed_rsi`, meaning local-first Pulse continuation, governed self-change rehearsal, exact evidence gates, and control-plane readback. - Unsupported claim: `full_autonomous_self_mutating_rsi`, unless future evidence proves production authority, claim-publish approval, and Covenant acceptance. - AO2 must not write its own permission slip. Blueprint/Covenant authorization and human/operator approval remain outside the mutation loop. ## Command Flow | Purpose | Command | Primary evidence | | --- | --- | --- | | Claim-readiness audit | `npm run rsi:claim-readiness` | `target/rsi-claim-readiness/latest/summary.json` | | Governed self-change dry run | `npm run rsi:self-change-dry-run` | `target/rsi-self-change-dry-run/latest/summary.json` | | Live rehearsal | `AO2_RSI_LIVE_SELF_CHANGE_REHEARSAL=1 npm run rsi:live-self-change-rehearsal` | `target/rsi-live-self-change-rehearsal/latest/summary.json` | | Rehearsal readback index | `npm run rsi:live-self-change-readback-index` | `target/rsi-live-self-change-readback-index/latest/summary.json` | | Improvement gate | `npm run rsi:improvement-evidence-gate` | `target/rsi-improvement-evidence-gate/latest/summary.json` | | Improvement trend | `npm run rsi:improvement-trend` | `target/rsi-improvement-trend/latest/summary.