/gov-metrics
Governance metrics — measure whether the gates actually work. Block rate, override/waiver rate, false-block proxy, R1-textual vs R2-mechanical share, and time-in-gate, computed from the verdict trail.
$ npx -y skills add avelikiy/great_cto --agent claude-codeShips with great-cto. Installing the plugin gets this command.
How it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/gov-metrics
Context preview
What this command does when you run it.
Governance metrics — measure whether the gates actually work. Block rate, override/waiver rate, false-block proxy, R1-textual vs R2-mechanical share, and time-in-gate, computed from the verdict trail.
Command definition
gov-metrics.mddescription: "Governance metrics — measure whether the gates actually work. Block rate, override/waiver rate, false-block proxy, R1-textual vs R2-mechanical share, and time-in-gate, computed from the verdict trail."
argument-hint: "[--since 30d] [--json]"
user-invocable: true
allowed-tools: Bash, Read
model: haiku
/gov-metrics — does the gate actually gate?
great_cto's moat is mechanical governance. A gate you can't measure is a vibe, not a control. This reports the numbers (idea adapted from SantanderAI/mech-gov-framework):
- **Block rate** — share of gate decisions that blocked.
- **Override / waiver rate** — blocks a human bypassed.
- **False-block proxy** — blocked-then-passed-unchanged (gate noise / over-firing).
- **R2 mechanical share** — verdicts enforced by CI/script (the moat) vs **R1 textual**
(a reviewer's prose judgment). High R1 share = gates lean on vibes, not enforcement.
- **Median time-in-gate** — minutes between consecutive gate decisions on a feature.
Run
PT="$(ls -d ~/.claude/plugins/cache/local/great_cto/*/ 2>/dev/null | sort -V | tail -1 | sed 's|/$||')"
[ -d "$PT" ] || PT="$(pwd)"
node "$PT/scripts/lib/gov-metrics.mjs" $ARGUMENTS
`--since 30d` limits the window; `--json` emits machine-readable output for the board.
Present
Show the CTO the table, then **one sentence of so-what**:
- R2 share < 40% → "gates lean on prose — move judgment into enforced checks."
- false-block proxy > 30% → "gates over-fire — calibrate severity (anti-inflation)."
- block rate ~0% over many decisions → "the gate never says no — is it real?"
Governance is the product. These numbers are how you prove it.
Read more
description: "Governance metrics — measure whether the gates actually work. Block rate, override/waiver rate, false-block proxy, R1-textual vs R2-mechanical share, and time-in-gate, computed from the verdict trail." argument-hint: "[--since 30d] [--json]" user-invocable: true allowed-tools: Bash, Read model: haiku
/gov-metrics — does the gate actually gate?
great_cto's moat is mechanical governance. A gate you can't measure is a vibe, not a control. This reports the numbers (idea adapted from SantanderAI/mech-gov-framework):
- **Block rate** — share of gate decisions that blocked.
- **Override / waiver rate** — blocks a human bypassed.
- **False-block proxy** — blocked-then-passed-unchanged (gate noise / over-firing).
- **R2 mechanical share** — verdicts enforced by CI/script (the moat) vs **R1 textual**
(a reviewer's prose judgment). High R1 share = gates lean on vibes, not enforcement.
- **Median time-in-gate** — minutes between consecutive gate decisions on a feature.
Run
PT="$(ls -d ~/.claude/plugins/cache/local/great_cto/*/ 2>/dev/null | sort -V | tail -1 | sed 's|/$||')" [ -d "$PT" ] || PT="$(pwd)" node "$PT/scripts/lib/gov-metrics.mjs" $ARGUMENTS
`--since 30d` limits the window; `--json` emits machine-readable output for the board.
Present
Show the CTO the table, then **one sentence of so-what**:
- R2 share < 40% → "gates lean on prose — move judgment into enforced checks."
- false-block proxy > 30% → "gates over-fire — calibrate severity (anti-inflation)."
- block rate ~0% over many decisions → "the gate never says no — is it real?"
Governance is the product. These numbers are how you prove it.
Don't buy software. Get the work done. GreatCTO ships AI autopilots that run a whole business function — medical coding, legal docs, procurement, accounting, IT, tax — from intake to outcome. A qualified human signs only the judgment calls. Live connectors, built-in compliance.
Repo: avelikiy/great_cto
Other commands on great-cto.
- /aedt-bias-audit
HR-AI / AEDT bias audit. Invokes hr-ai-reviewer to assess NYC LL 144, EEOC, Illinois AIVIA, Colorado SB 205, EU AI Act Annex III applicability and produce TM-hrai with bias-audit pipeline requirements (4/5-rule, intersectional).
Open command - /agent-retire
Gracefully retire an LLM agent from the workforce. Archives prompt, removes from sync list, keeps verdicts for audit. Like firing a human — but reversible.
Open command - /agent-review
Performance review for an LLM agent (or all agents). Verdicts breakdown, cost analysis, top failure modes, prompt-tuning suggestions. Like a human '1:1' but for AI workforce.
Open command - /api-contract-review
API platform contract review. Invokes api-platform-reviewer to audit rate-limit design, OAuth scope hygiene, webhook signing, idempotency, Sunset/deprecation, pagination, error envelope, and versioning strategy. Critical before v1 GA.
Open command - /audit
Audit an existing codebase. Detects stack, finds gaps, creates tasks, generates PROJECT.md.
Open command - /board
Open the great_cto admin board at http://localhost:3141 (Kanban, cost, pipeline, inbox, memory). Starts it in background if not running.
Open command