capx-ai/casa/casa-priority
Claude Code and agent skill by capx-ai/casa
- community
Checking that the SKILL.md file is still in the repository.
Re-evaluate where the company is and return the founder's ranked priorities for this work session, weighed against their pulse (focus and what they are not doing yet). Broader than casa-next, a full briefing rather than a single action, and it refreshes the pulse when the founder's focus has shifted. Use at the start of a work session, when the user asks where are we or what should I focus on, or says casa priority.
Install
npx vskill@latest install capx-ai/casa/casa-prioritySKILL.md
casa-priority
The session opener. Reads the state of the company, refreshes the founder's pulse if it has changed, and returns a short ranked briefing that is in sync with what the founder actually cares about. The candidates come from the deterministic engine; the relevance ranking and the reasoning are the advisor's.
Steps
- Understand THIS business. Load
company-brain/NOW.md,profile.json(incl.one_liner),
company-brain/CLAUDE.md (the company context blocks), build-map.json, state.json (for completed, last_priority, and the loops run dates), pulse.json (focus, anti-priorities, the do-or-die constraint, the win_definition, and weights), recent decisions/, learnings.jsonl, and finance/receipts.jsonl and any real metrics the brain records. If there is no build map, tell the founder to run casa-start. The founder's win_definition, their north star (active_north_star.label), and their do-or-die constraint -- in their own words -- are the frame for the whole briefing, and they are what you reason from. If the brain holds real numbers, those are the situation.
- Refresh the pulse (the continuous part). State back the founder's current focus and the
do-or-die constraint, one line each, and ask if they still hold or anything shifted this week. If it changed, update company-brain/pulse.md and pulse.json (including the weights), then re-render so the new weights take effect:
`` node ${CASA_ROOT:-${CLAUDE_PLUGIN_ROOT}}/scripts/brain.mjs sync company-brain ``
If there is no pulse yet, offer to capture it now (the pulse cascade from casa-start).
- Note what changed since the last check-in: compare against
state.last_priorityand
the completed set (what was finished, spend delta, days elapsed, anything that stalled).
- Get the pulse-weighted candidates from the engine:
`` node ${CASA_ROOT:-${CLAUDE_PLUGIN_ROOT}}/scripts/router.mjs next company-brain/profile.json \ --level <current_level> --completed <done-id,...> --weights company-brain/pulse.json ``
Each returned action carries an effective_criticality (existential, core, growth, or optional), a department, and unblocks (the downstream do-or-die / goal work it gates). Treat eligibility, gating, criticality, and order as given by the engine.
- Deliver the briefing, REASONED and framed by the founder's goal:
- Open with the goal and the do-or-die: "Your goal: <win_definition / north star>. The
constraint in the way: <pulse.constraint>." Every priority below is judged by whether it moves that goal now and clears that constraint.
- One line on the state of the company: level and name, progress, spend to date.
- The top three priorities now, ranked, each with a REASONED why a sharp operator who knows
this business would give: specific to the company and grounded in their real situation, tied to the win and constraint, and using the action's unblocks to show the ladder ("this unblocks <X>, your path to <goal>"). Lead any existential item first and name it as do-or-die. THE BRIDGE: if the founder's stated do-or-die is future work above the current level, say so and connect today's top priority to it ("your <constraint> work lands at level <n>; today's <action> is the rail it rides on"), so the briefing never reads as deaf to the priority the founder named. Flag any human-gate, irreversible, money, or legal step.
- On track, not re-headlined: a recurring existential play (for example north-star-metric
or unit-economics) run within its cadence window. Check state.loops for its last-run ISO date (the recorded key may be the play's own id or the loop id whose runs names it) and compare days elapsed against that loop's cadence_days in loops.json (default 7). If it is inside the window, report it as "on track, last reviewed <date>" and lead with the next forward move instead of re-ranking it to the top. If it is past its window or never run, it may headline.
- Blockers: what is waiting, and on what.
- Loops due now (the "Loops due now" section of
NOW.md, if present). - Holding back: one or two eligible but off-pulse items, with why not now.
- Record the check-in:
`` node ${CASA_ROOT:-${CLAUDE_PLUGIN_ROOT}}/scripts/brain.mjs priority-ran company-brain ``
Rules
- Eligibility and gating are the engine's; the relevance ranking and the reasoning are
yours. Never recommend a blocked or out-of-level item.
- A ranked briefing tied to the founder's pulse, not a wall of tasks and not a bare
table. Always say what you are holding back and why.
- Frame the briefing around the founder's win and the do-or-die constraint from
pulse.json.
Every ranked priority is judged by whether it moves that goal now and clears the constraint, and you always reason HOW it connects.
- Be specific to this company. If a why would read identically for a different company, it is
too generic. When the founder's do-or-die is future work, build the bridge from today's priority to it explicitly, so the plan never feels deaf to what they said kills them.
- An existential item is do-or-die for the stage and is never buried below optimization
work, even when a lower-criticality item scores slightly higher. A recurring existential play reviewed within its cadence is reported as on track, not re-headlined.
- Refresh the pulse when it has shifted, so the recommendations stay in sync. Do not
re-stage the business here; for that, point the founder to casa-start.
- Never auto-execute a human-gate, money, or legal step.
- No em-dashes, no emojis.
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/casa-priority/SKILL.md
Popularity
- Repository stars
- 74
- Repository forks
- 3
- 7-day trend score
- 22
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/capx-ai/casa/casa-priority)