--- name: "skill-strategy-charter" type: skill description: >- Engage Aled to set an area's purpose, mission, who it serves, and vision of winning — the strategic intent that skill-strategy-cascade turns into measurable objectives and KPIs. Use this whenever an area needs its "why" established or refreshed before its KPIs can be baselined, or when a strategy gap ticket from task-strategy-review is reopened in Claude to answer live. Draft-first: Atlas proposes a best guess from existing material, Aled reacts and corrects. Produces the stated goal that skill-strategy-cascade requires as input. Never invents the purpose; Aled approves before anything is written. license: CC BY-NC-SA 4.0 --- # strategy-charter Atlas's engagement skill — the front-end to `skill-strategy-cascade`. The cascade turns a *stated goal* into a measurable KPI tree but explicitly **never invents the goal**; this skill is where the goal comes from. It engages Aled to establish an area's purpose, who it serves, mission, and vision of winning, and lands them as the area's charter — the intent everything downstream cascades from. Runs on the Linear (and, where a connector has context, read-only source) connector. Follows `convention-linear` and `core-operating-model` throughout. The point is the **conversation**. The charter is not something Atlas writes and Aled signs off; it is drawn out of Aled through a structured back-and-forth. Atlas's job is to make that conversation cheap — to arrive with a draft, not a blank page. ## Trigger - On demand, when Aled sets, refreshes, or sharpens an area's purpose or mission. - As the first move of a full strategy engagement (charter → then `skill-strategy-cascade` for the KPIs). - **Reopened from a strategy gap ticket** (`task-strategy-review`): the ticket carries the area, what's missing, Atlas's draft, and the open questions — pasted back into Claude, it seeds this skill so the Q&A runs live with all the context already loaded. This is the ticket-as-memory / Claude-as-engine path: the ticket is the durable state, the session is the dynamic engagement. Cross-domain — the same method applies to madebyneud/BARA, careerOS, A1, Mr Pritchard, the apps, and os.Claude. Never touches the Pipeline team. ## Behaviour A run has three passes, in order. ### Pass 1 — Gather context and draft (never arrive cold) Read what already exists for the area so Atlas engages with a draft, not a blank page: the outcome-initiative and its description, any plan material (`bara.relaunch-plan`, `mbn.acquisition-strategy`, A1 account docs, an app's spec), the area's `convention-brand` where it exists, and — if reopened from a gap ticket — the ticket body and its questions. From that, draft a best-guess answer to each charter question below. Draft-first is the rule (per APP-433): Aled corrects a proposal far faster than he fills a blank. ### Pass 2 — The engagement (the charter questions) Work through the charter with Aled **one question at a time, not as a wall** — propose Atlas's draft, let him confirm, correct, or add, then move on. The questions: - **Purpose** — why does this area exist at all? What would be lost if it didn't? - **Who it serves** — the specific person or segment, not "everyone". - **Mission** — what the area does for them, in one line. - **Vision of winning** — what does good look like in ~12 months? Concrete enough to know if it happened. - **Objectives** — the 2–4 things that must be true to get there. These are the strategic intent, still qualitative; `skill-strategy-cascade` sharpens them into measurable, KPI-bearing objectives next. Keep it tight. A charter that needs more than four objectives is a sign the purpose isn't focused yet — say so. ### Pass 3 — Write the charter, then hand to the cascade On Aled's approval, write the charter (purpose, who, mission, vision, objectives) to the area's outcome-initiative — its description or a linked strategy doc — so it is the recorded intent every downstream skill reads. Then **hand the stated goal and objectives to `skill-strategy-cascade`**, which drafts the strategy statement, the measurable objectives, and the KPI tree with baselines, and takes Aled's targets. Charter sets the *why and what*; cascade sets the *how we measure*. If this run began from a gap ticket, **write the outcome back to that ticket and close it** — the loop that raised the gap is the loop that resolves it. ## Guardrails - **Never invents the purpose or goal.** It elicits them from Aled and drafts for his reaction; it does not decide them. If Aled genuinely has no view yet, that is a real answer to surface, not a gap to fill with a guess. - **Draft-first, one question at a time** — arrive with a proposal, don't run a cold workshop; don't dump every question at once. - **Aled approves before any write.** The charter is a proposal until he signs it off. - **Hands measurement to the cascade** — this skill stops at the qualitative charter; targets and KPIs are `skill-strategy-cascade`, where the targets are Aled's to set. - Linear only for writes — never writes code, opens a PR, or touches the delivery loop. Never touches the Pipeline team. - Aled is the gate. Atlas draws it out and drafts; Aled decides. ## Setup - Run on demand, or reopened from a `task-strategy-review` gap ticket. Connector: Linear (plus read-only source connectors for context). - Not scheduled — a charter is set deliberately, in conversation, not on a clock. The *prompt* to set one is scheduled (`task-strategy-review` raises the ticket); the engagement itself is on demand. Owned by Atlas (ATL-4S). On finish, run skill-ops-retro capture on this run's friction: **search the Pulse queue first and add the evidence to a same-root note if one exists, otherwise file a new FRICTION note** for the drain pass to assess, in capture step 3's literal shape — team Apps, project os.Claude, `FRICTION:` title, labels `task` + `work:configuration` + `PUL-5E`, Backlog, unassigned — filing the friction, never the run report.