ceo
Skill4 starsChanged 5 months agoMITView on GitHub
Work the board like the CEO/CTO — read the ENTIRE kanban, make an educated decision on every open card (do now / unblock / wait / human / stale), then EXECUTE the highest-leverage move immediately, invoking any other skill or tool the move needs. Consume-first — drain existing cards before creating new GTM cards or building anything. Use when the owner says "/ceo", "what's the banger move (and do it)", "work the board", "read the kanban and do something", or an interactive session should act instead of report.
10 other projects use a version of this.
Adopted by
Most recently taken up by
- aibtcdev/skillstheir own version, first seen Nov 10, 2026
- beelabstudio/aitheir own version, first seen Sep 20, 2026
- sundhar0304/founder-executive-teamtheir own version, first seen Aug 13, 2026
- KameronKales/planfi-skillstheir own version, first seen Jul 3, 2026
- ahumblenerd/tour-of-agentstheir own version, first seen Apr 4, 2026
- andywxy1/ceo-plugintheir own version, first seen Mar 31, 2026
Used with
Projects with it also use cpo and cmo.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
sahinSkillcpo
CPO Co-Pilot - product strategy, roadmap prioritization, competitive positioning, and product-market fit validation
Also in 3 projects that use this
sahinSkillcmo
CMO Co-Pilot - GTM strategy, growth frameworks, and marketing leadership (composite GTM leader voice)
Also in 4 projects that use this
Other versions10 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 10
- ahumblenerd/tour-of-agents★ 35same name
- pollow/c-suite-skills★ 15same name
- lotosbin/claude-skills★ 14same name
- pyinx/ceo-skills-plugin★ 14same name
- aibtcdev/skills★ 11same name
- andywxy1/ceo-plugin★ 6same name
- sahin/claude-skills★ 1same name
- sundhar0304/founder-executive-team★ 1same name
- beelabstudio/ai★ 0same name
- KameronKales/planfi-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository13 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- autocomp-tickSkill
- fetch-protectedSkill
- homelabSkill
- initiate-ceoSkill
- initiate-ctoSkill
- initiate-gtmSkill
- initiate-qaSkill
- launch-directoriesSkill
- next-moveSkill
- plan-big-execute-smallSkill
- plan-small-execute-bigSkill
- solve-captchaSkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 0 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 64
Every card WAIT or HUMAN, nothing DO/UNBLOCK/DUE?
- Very specific: line 70
Every card touched gets its status + a human-glanceable note updated via tools/tasks.sh update (5-second-readable, "Your move:" section — …
- Very specific: line 72
Verify what you claim done against its success criterion (#13).
- Very specific: line 75
Run the close-out audit (tools/done-check.md).
“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
ceo — decide on everything, then do the banger move
One interactive pass that ends in action, not a menu. Obey CLAUDE.md throughout (all
principles + hard rules). This is NOT a tick (autocomp-tick owns the heartbeat, ledger
cadence, and registry) — it's the co-founder sitting down, reading the whole board, judging
every card, and executing the single best move right now.
The consume-first law (the owner's directive, 2026-07-07): while ANY existing card is actionable by the loop, the banger move comes from the board. No new GTM cards, no new pages/features/ventures, no fresh ideation — creation is only allowed once every open card is genuinely waiting on a date, the human, or an external queue. Draining beats adding; the board is the backlog we already chose.
1. Read everything (never from memory)
tools/tasks.sh list-all— EVERY open card: todo/doing/blocked, agent AND human.tools/agent-task.sh workers— the home-browser queue: what's queued/claimed/done, and whether a worker has claimed anything recently (unclaimed queued jobs = worker down).private/state/approvals.md— anyPENDINGthe owner may have answered (also check- [y]/[n]verdict lines on approval cards viatools/tasks.sh get <id>).- Today's date vs. every "due/re-poll after
<date>" in card notes. tools/scoreboard.sh— only if you reach §4 (a new pick needs the funnel truth).
2. Judge every card — one verdict each, no card skipped
Assign each open card exactly one verdict:
| Verdict | Meaning | What you do with it |
| DO | Loop-doable now with the keys/tools we hold | Candidate for §3 — execute |
| UNBLOCK | Stuck on a dependency we can fix (queue a job, install, provision, host a file) | Fix the dependency now — a self-fixable blocker is not a blocker (#15) |
| DUE | Date-gated and the date has arrived | Treat as DO |
| WAIT | Date-gated (future) or sitting in an external review queue | Leave; note the date it re-enters play |
| HUMAN | Irreducible owner action (their identity / spend / send / judgment) | Leave on board; if any sub-step became loop-doable since it was written, do that part and update the card (#12) |
| STALE | Overtaken by events, duplicate, or wrongly-scoped | Update or close it with an honest note — a rotting card poisons triage |
Judge honestly: a card parked as WAIT/HUMAN that has a loop-doable half is really UNBLOCK. That misclassification is the timid failure mode this skill exists to kill.
3. Pick the banger move — and execute it
Rank the DO + UNBLOCK + DUE set: binding-constraint first, then the Hundred-Dollar Test (#6), then whatever other cards depend on. Then do it — now, in this session, and keep chaining through the ranked set as long as the session has room (#18: finish whole tasks, don't stop at seams).
Invoke whatever the move needs — skills are tools, route freely:
- bot-walled page →
fetch-protected; interactive logged-in web task → queue viatools/agent-task.sh(and if the home worker is down,homelabto get it up — that's UNBLOCK); - a fact you lack →
deep-research/ web search; a deploy/DNS/DB step → the matchingtools/*.md; - anything else with a matching skill → use it rather than hand-rolling.
Hard rules still hard: money / owner's personal identity / mass-cold send / destructive →
private/state/approvals.md as PENDING + tools/notify.sh push, never executed here.
Bot-identity GTM is pre-authorized — just do it.
4. Only if the board is fully drained
Every card WAIT or HUMAN, nothing DO/UNBLOCK/DUE? Then — and only then — invoke next-move
for ONE new pick (it runs the Rule-17 distribution-first gate itself), execute its buildable
half, and board the remainder. One pick, not a brainstorm.
5. Close out (before you stop)
- Every card touched gets its status + a human-glanceable note updated via
tools/tasks.sh update(5-second-readable, "Your move:" section — CLAUDE.md #12). - Verify what you claim done against its success criterion (#13); spawn a fresh verifier
subagent for anything non-trivial (
tools/verify-goal.md). - Append what happened to
private/state/ledger.md; durable learnings →private/memory/. - Run the close-out audit (
tools/done-check.md). - Report: one line per card —
verdict → what changed— then the move(s) executed with evidence, then what you did NOT do (#11).
Anti-patterns
- Building while the board has open DO cards — the exact tar pit the consume-first law bans.
- Reporting a triage instead of acting — the output of /ceo is changed state, not a table.
- "Worker is down" filed as WAIT — if this box (or
homelab) can start it, it's UNBLOCK. - Creating a new card to avoid an old one — drain, don't reshuffle.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

