plan
Skill0 starsChanged 3 months agoMITView on GitHub
Scope any idea, initiative, or "let's build X" through the readiness gates into an executable plan on one durable file. Run when the user says "what's the plan for...", "scope this out", "I want to launch...", or "how should we build X" after the founder has decided to proceed. A "whether" question belongs to the gauntlet. Also run when the founder hands over ANY build/execute request whose initiative is not yet planned. This skill is the gate that intercepts raw "go, build it" asks.
184 other projects use a version of this.
Adopted by
Most recently taken up by
- event-driven-io/emmetttheir own version, first seen Nov 16, 2026
- champ18ion/claud18codetheir own version, first seen Nov 16, 2026
- FAHTHIZ/oasistheir own version, first seen Nov 16, 2026
- chuckplayer/claude-agent-packtheir own version, first seen Nov 16, 2026
- ansvisor/ansvisortheir own version, first seen Nov 15, 2026
- timmo001/skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use context-optimize, heartmula and 10 more.
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.
SireJeffSkillcontext-optimize
Context Optimization - Orchestrate RPI workflow TODO generation with interactive scoping
Also in 4 projects that use this
math-incSkillheartmula
Set up and run HeartMuLa, the open-source music generation model family (Suno-like). Generates full songs from lyrics + tags with multilingual support.
Also in 9 projects that use this
Tommy-ywSkillwebhook-subscriptions
Create and manage webhook subscriptions for event-driven agent activation, or for direct push notifications (zero LLM cost). Use when the user wants external services to trigger agent runs OR push notifications to chats.
Also in 5 projects that use this
math-incSkillminecraft-modpack-server
Set up a modded Minecraft server from a CurseForge/Modrinth server pack zip. Covers NeoForge/Forge install, Java version, JVM tuning, firewall, LAN config, backups, and launch scripts.
Also in 5 projects that use this
john-data-chenSkilltouchdesigner-mcp
Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.
Also in 6 projects that use this
gaebalaiSkillcodex-system
PROACTIVELY consult Codex CLI, your highly capable supporter with exceptional reasoning and task completion abilities. Codex is a trusted expert you should ALWAYS consult BEFORE making decisions on: design choices, implementation approaches, debugging strategies, refactoring plans, or any non-trivial problem. When uncertain, consult Codex. Don't hesitate - Codex provides better analysis. Explicit triggers: "think deeper", "analyze", "second opinion", "consult codex".
Also in 3 projects that use this
- +3 more from math-inc
Other versions184 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 184
- Yeachan-Heo/oh-my-codex★ 34ksame name
- crbnos/carbon★ 2.8ksame name
- graphistry/pygraphistry★ 2.6ksame name
- zereight/gitlab-mcp★ 2.1ksame name
- math-inc/OpenGauss★ 1.3ksame name
- kdlbs/kandev★ 1.1ksame name
- suboss87/FDEOps★ 940same name
- codewithmukesh/dotnet-claude-kit★ 808same name
- witnesstodark/mr-mak-workspace★ 800same name
- loongclaw-ai/loongclaw★ 640same name
174 more not listed
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.
The file
Show its rules and text
We split this skill into the rules it gives an agent
16 rules. 0 are common, 6 are distinctive, and 10 are very specific to this project.
0
Common
Most of the 168 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
10
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Every initiative gets ONE file that travels its whole life:
- Very specific: line 25
Before any step, read references/CONVENTIONS-core.md.
- Very specific: line 31
Run the shared contract's Universal preflight.
- Very specific: line 38
Read pending queue.md entries addressed to plan.
- Distinctive: line 54
Check the stages in order, and stop at the first one missing:
- Very specific: line 61
BIG bet, state
survived: read the latest Gauntlet verdict. - Very specific: line 65
Create at most an idea stub.
- Distinctive: line 77
State is unknown or malformed: stop for repair. Never infer the nearest legal state.
- Distinctive: line 97
Every line must be judgeable without homework.
- Very specific: line 100
Run the shared durable claim admission gate before this checkpoint and again before filing.
- Distinctive: line 110
Ask before you assert: never present a quantity, duration, sample size, date, or causal motive in prose before asking the founder for it or …
- Very specific: line 121
STOP. Read references/initiative-template.md now: it holds the file's exact section shape and the roadmap rules.
- Distinctive: line 130
Write decisions with their rationale, never task spam.
- Very specific: line 140
Every initiative advancing beyond idea carries both checked references in frontmatter: internal_receipt: <path>#<quote-anchor> to business …
- Very specific: line 161
When this plan consumes a queue entry, mark that same ID done with the initiative path only after the initiative legally reaches planned.
- Very specific: line 167
Run scripts/graph-audit on every touched durable file.
“Common” is measured: kept by at least 6 in 10 of the 168 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
plan: scope it properly, or don't build it at all
Every initiative gets ONE file that travels its whole life:
idea ──> survived ──> planned ──> executing ──> reviewed ──> banked
^ ^
the gauntlet THIS SKILLPlan owns the planned arrow. Work that reaches sprint without passing through here does not
run. That refusal is the whole point of the collection, and this skill is where it lives.
Before any step, read references/CONVENTIONS-core.md. This resolves from the
installed skill directory to the plugin's shared contract; the founder's working folder is
never used to locate it. Its state, receipts, filing, and dedupe rules govern this workflow.
Step 0: Load the ground
Run the shared contract's Universal preflight. A Fresh, Partial, or Wrong-folder result stops and runs the co-founder-setup skill before anything else (re-sync for Partial), without creating an initiative or vision
stub. On Ready, read the charter (AGENTS.md / CLAUDE.md), then check the charter-mapped vision
path immediately. If the canonical vision is missing, say the
founder is planning against nothing, route to vision, and stop without creating or sizing an
initiative, changing a queue entry, or writing a log row. Only after vision exists, consult the
vault: what does
wiki/ already know about this idea, this customer, past attempts? Cite what you find by file.
Read pending queue.md entries addressed to plan; an offer queue item points to the canonical
wiki/business/offer.md rather than carrying copied terms. Then locate the initiative: an
existing file in wiki/initiatives/, or a new one.
Done when: you hold the charter's laws, the vault's relevant history, and the initiative file.
Step 1: Size the stakes
One judgment, made out loud: SMALL (hours, reversible, no real money) or BIG (weeks, money, customer-facing, hard to undo). Uncertain resolves to BIG. The size sets everything downstream: question count, research depth, whether the gauntlet is required.
Done when: the size is stated in one line with its reason, in the initiative file.
Step 2: The gate check
Check the stages in order, and stop at the first one missing:
- BIG bet, state still
idea: route to the gauntlet first. One sentence of why ("this is weeks of your life, it gets interrogated before it gets planned"), then the queue item contract: leave the initiative as anideastub, tell the founder to run the gauntlet on it, and write nothing further. The gauntlet alone flipssurvived; plan resumes after. A SMALL bet may proceed directly, recorded as: "Sized small, gauntlet waived." - BIG bet, state
survived: read the latest Gauntlet verdict. Only SURVIVES with no Parked recommendation may proceed; DIES or PARKED routes back to the named evidence gap without a state change. - The ask is raw execution ("just build it", "go"): refuse the build, start the scope.
Create at most an
ideastub; the missing stage isplanned, and the refusal shape from the shared contract carries the message: name the missing stage, ask the first missing question. Interviewed, never scolded. - State is already
planned: this is replanning, not a new transition. Preserve the state, run the checkpoint on the changed scope, and update the roadmap without erasing history. - State is
executing: a same-destination amendment preserves checked steps and staysexecuting. A changed destination creates a separateideawhose body links this initiative;reopensstays null because the predecessor is still active. With the founder's confirmation, move every unchecked old step toRoadmap → Stoppedwith today's date and the reason. Review records the stopped outcome on the old arc; the new idea earns its own gates. - State is
reviewedorbanked: the arc is closed. New work gets a new linkedidea; the closed file stays unchanged. - State is unknown or malformed: stop for repair. Never infer the nearest legal state.
Done when: every required stage before planned is genuinely done, or the founder has been
routed to the one that is missing.
Step 3: Scope the WHAT
One question per turn, recommended answer attached, founder's words captured. Facts you can get yourself (from the vault, from research), get yourself. Probe only where a real gap shows: what is the smallest version that still delivers, and what does the current workaround cost? A concrete founder earns zero probes.
Then the checkpoint, before the Scope and Roadmap sections or any state change are written
(the idea stub and its sizing line may already exist). Present on one screen:
- What we're doing (2-3 lines in their words)
- Key tradeoffs (the calls made and why)
- Out of scope (what this deliberately is not)
- Needs your call (anything only they can rule on)
Every line must be judgeable without homework. Any correction: integrate, re-present the whole checkpoint, wait for the yes again. A revision is never a confirmation.
Run the shared durable claim admission gate before this checkpoint and again before filing. Each
number, customer fact, and causal statement carries a recomputation, an exact checked quote from
an open source file, or Founder-stated: unverified. Earlier gauntlet or offer prose is not trusted
as a receipt merely because it is already durable.
Apply the shared parameter gate to durations, counts, caps, sequence assumptions, and completion thresholds. Ask one founder question when a missing value blocks the scope; otherwise keep it as a working hypothesis under Needs your call or Parked recommendations. Imported unverified offer or gauntlet values never become roadmap commitments.
Ask before you assert: never present a quantity, duration, sample size, date, or causal motive in prose before asking the founder for it or citing a receipt for it.
A wiki/decisions/** page is a T1 consent surface (shared contract, Immutable consent receipts):
plan may create one only with the founder's [receipt: raw/inbox/<file>#<exact words>] for the
exact call. Without that yes, record the tradeoff in Key decisions and write no decision page.
Done when: the founder has said yes to the exact scope about to be written.
Step 4: Plan the HOW
STOP. Read references/initiative-template.md now: it holds the file's exact section shape and
the roadmap rules. An improvised structure breaks sprint's ability to execute it.
Snapshot every pre-existing prior-stage section before editing. Patch Scope, Roadmap, Needs your
call, and Parked recommendations in place; preserve the complete Gauntlet verdict, advisor board,
objections, exact receipt lines, and append-only Log. Scope links to [Gauntlet verdict](#gauntlet-verdict).
log.md points to the event and never stores the board. Compare the preserved sections byte for
byte before the state flip; any loss aborts the transition.
Write decisions with their rationale, never task spam. Every consequential recommendation in
Key decisions carries both receipt legs from the shared contract: a vault fact and a checked
external source. A missing leg moves the item to Parked recommendations in the contract's
canonical shape; it never moves to "Needs your call" because evidence gathering is not a founder
judgment. The
roadmap section follows fog-of-war
discipline: chart in detail only what is visible now as session-sized steps, keep the rest in
"Not yet decided" where it graduates as the frontier advances, and keep "Out of scope" as the
list that never graduates.
Every initiative advancing beyond idea carries both checked references in frontmatter:
internal_receipt: <path>#<quote-anchor> to business evidence outside initiative/content output,
and external_receipt: wiki/research/<file>.md#<quote-anchor> to dated checked research. This is a
state invariant, independent of wording or size. Missing fields keep the initiative PARKED and
below planned.
Done when: the initiative file matches the template with zero empty sections (an honest "Not yet decided" or "Parked recommendations: None." entry counts as content).
Step 5: Flip the state and hand off
Re-read the current state immediately before writing. Set state: planned only for the legal
idea → planned SMALL waiver or survived → planned transition. Replanning stays planned; a
same-destination executing amendment stays executing. Any other state follows Step 2 instead
of moving backward. Bump today's date and append the log row:
## [date] plan | <initiative title>. Ensure index.md and hub/initiatives.md point to the
initiative under the bucket matching its current state, update the index roadmap summary and
baton.md, and reconcile the consumed queue entry as one linked-state transaction. Then close
with exactly two lines: what got decided
(one line), and the ONE next command: "say 'let's work' and sprint takes the first step."
When this plan consumes a queue entry, mark that same ID done with the initiative path only
after the initiative legally reaches planned. A route to vision, gauntlet, or receipts leaves the
entry pending.
Mandatory filing gate
- Run
scripts/graph-auditon every touched durable file. - Fix every violation and rerun it. The plan, state transition, log, and queue completion remain incomplete until it exits zero.
Done when: the legal state is preserved or advanced, the log row exists, and the founder knows the next move.
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.

