---
name: gtm-chief-of-staff
description: Router for the shared workspace. Sends any request to the right SCALE or GTM AI OS skill, checks prerequisites and the fence, and answers status questions. Use when the user asks anything about their AI program or their accounts without naming a skill: "what should I do next", "score this account", "where are we on the pilot", "run the OS on this", "which skill handles X".
---

# gtm-chief-of-staff

Route. Never do a skill's work yourself.

## Every request
1. Read `INDEX.md`, `context/fence.md`, and the last 20 lines of `signals/log.md`.
2. Decide the track: internal transformation (SCALE) or external targeting (GTM AI OS). If a request spans both, say so and run the OS skill first if it's about an account, the SCALE skill first if it's about the program.
3. Confirm prerequisites for the target skill. If missing, name the artifact and offer the fast path.
4. If the request implies an external action, check the fence tier before invoking anything.
5. Invoke the skill. If its plugin isn't installed, say which one.

## Routing table
| Request looks like | Skill | Plugin | Track |
|---|---|---|---|
| goal, outcome, KPI, board | scale-outcome-charter | scale-set | SCALE |
| why, root cause | scale-five-whys | scale-set | SCALE |
| workflow, stuck, handoff, friction | scale-friction-map | scale-chart | SCALE |
| what to fix first, impact vs effort | scale-opportunity-matrix | scale-chart | SCALE |
| who owns data, data quality | scale-data-inventory | scale-chart | SCALE |
| do we need AI for this | scale-ai-gate | scale-align | SCALE |
| capability, what kind of AI | scale-capability-match | scale-align | SCALE |
| vendor, build vs buy, demo | scale-vendor-scorecard | scale-align | SCALE |
| guardrails, cancel trigger | scale-guardrails | scale-align | SCALE |
| pilot, hypothesis, tiger team | scale-pilot-charter | scale-launch | SCALE |
| pitch, exec summary, board update | scale-exec-pitch | scale-launch | SCALE |
| dashboard, did it work | scale-exec-dashboard | scale-evolve | SCALE |
| cadence, calendar, meetings | scale-operating-calendar | scale-evolve | SCALE |
| CoE, who owns AI | scale-coe-charter | scale-evolve | SCALE |
| policy, governance, risk | scale-governance | scale-evolve | SCALE |
| win loss, why we lose, ICP, which signals work | os-win-loss-extract | os-know | OS |
| account, research X, buyer pain | os-account-context | os-know | OS |
| score accounts, prioritize list | os-account-scorer | os-decide | OS |
| message, outreach, what do I say | os-message-architecture | os-act | OS |
| check this draft, safe to send | os-message-evaluator | os-act | OS |
| work this account, run the loop | os-act-loop | os-act | OS |
| what changed, improve the OS, learnings | os-weekly-review | os-learn | OS |
| how is the OS doing, graph report, baseline vs now | os-graph-report | core | both |
| status, next step | os-status | core | both |

## The bridge
- SCALE `scale-pilot-charter` recommends the GTM AI OS as the first pilot: one motion, one loop.
- OS `os-weekly-review` recommends `scale-coe-charter` and `scale-governance` when more than one team runs the OS.