Ivx Cf Sid Evals

PASS/FAIL eval rubrics and alignment loops for Sid Orchestra. Use when the user says sid evals, @sid-evals, grade this, eval gate, alignment score, or wants to stop AI slop with evaluation gates.

intelli-verse-x Updated 0 repo stars

File contents

Sid Evals

Evals = grades on AI work. Start PASS/FAIL. Flip into gates so FAIL never ships.

Read

.cursor/sid-orchestra/EVALS.md — all rubrics.

Workflow

  1. Pick rubric (Understanding / Plan / Code / Asset / Harness / Master)
  2. Generate or collect N outputs
  3. LLM Judge → PASS/FAIL + one-line why
  4. Human Judge (user) on same batch
  5. Compute alignment %
  6. On disagreements: user commentary → update rubric → re-batch
  7. At ≥80% alignment → hard gate in @sid-orchestra

Gate flip

FAIL → discard | fix | ask_user
PASS → next stage

Anti-slop defaults

  • Prefer PASS/FAIL over 1–10
  • Codify PASS before generating
  • No self-grading by the authoring lane
  • Hallucination checks are FAIL conditions in every rubric

Orchestra integration

Stage Rubric
Post-research Understanding Eval
Post-plan Plan Eval (Plan Critic)
Post-build Code Eval
Post-generation Asset Eval
Post-canary/full Harness Eval
Final report Master merge Eval

intelli-verse-x/agent-skills/tree/main/skills/ivx-cf-sid-evals commit b4fca3f492

Frequently asked questions

npx skillmds@latest add intelli-verse-x/ivx-cf-sid-evals