Wayfinder

A loose idea has arrived — too big for one agent session, and wrapped in fog: the way from here to the destination isn't visible yet. Wayfinding is about finding that way, not charging at the destination. This skill charts the way as a shared map on the repo's issue tracker, then works its decision tickets — questions whose resolution is a decision, not slices of a build to execute — one at a time until the route is clear.

The destination varies per effort, and naming it is the first act of charting — it shapes every ticket. It might be a spec to hand off and iterate on, a decision to lock before planning starts, or a change made in place like a data-structure migration. The map is domain-agnostic — engineering work, course content, whatever fits the shape.

Plan, don't do

Wayfinder is planning by default: each ticket resolves a decision, and the map is done when the way is clear — nothing left to decide before someone goes and does the thing. The pull to just do the work is usually the signal you've reached the edge of the map and it's time to hand off. An effort can override this in its Notes — carrying execution into the map itself — but absent that, produce decisions, not deliverables.

Refer by name

Every map and ticket is an issue, so it has a name — its title. In everything the human reads — narration, the map's Decisions-so-far — refer to it by that name, never by a bare id, number, or slug. A wall of #42, #43, #44 is illegible; names read at a glance. The id and URL don't vanish — a name wraps its link — but they ride inside the name, never stand in for it.

The Map

The map is a single issue on this repo's issue tracker, labelled wayfinder:map — the canonical artifact. Its tickets are child issues of the map.

The map is an index, not a store. It lists the decisions made and points at the tickets that hold their detail; a decision lives in exactly one place — its ticket — so the map never restates it, only gists it and links.

Where the map, its child tickets, blocking, and frontier queries physically live is tracker-specific. Read TRACKER.md beside the repo's engineer skill — its "Wayfinding operations" section says how this repo expresses child issues, blocking edges, and the frontier query. If there is no TRACKER.md, resolve the tracker for this run from what the user gave you, the git remote, and whichever CLI is installed and authenticated — and default to the local-markdown tracker when nothing resolves.

The engineer skill's directory varies by harness, so find it by glob:

ls skills/*-engineer/TRACKER.md */skills/*-engineer/TRACKER.md .*/skills/*-engineer/TRACKER.md 2>/dev/null

The map body

The whole map at low resolution, loaded once per session. Open tickets are not listed — they are open child issues, found by query.

## Destination

<what reaching the end of this map looks like — the spec, decision, or change this effort is finding its way to. One or two lines; every session orients to it before choosing a ticket.>

## Notes

<domain; skills every session should consult; standing preferences for this effort>

## Decisions so far

<!-- the index — one line per closed ticket: enough to judge relevance, then zoom the link for the detail the ticket holds -->

- [<closed ticket title>](link) — <one-line gist of the answer>

## Not yet specified

<!-- see "Fog of war": in-scope fog you can't ticket yet; graduates as the frontier advances -->

## Out of scope

<!-- see "Out of scope": work ruled beyond the destination; closed, never graduates -->

Tickets

Each ticket is a child issue of the map; the tracker's issue id is its identity. Its body is the question, sized to one 100K token agent session:

## Question

<the decision or investigation this ticket resolves>

Each ticket carries a wayfinder:<type> label — one of research, prototype, grilling, task (see Ticket Types).

The label fixes the HITL/AFK axis for three of those types: research is always AFK, prototype and grilling always HITL. task is the exception — it goes either way, so a task ticket states its axis on a **HITL** or **AFK** line under the question. Without it the axis is unrecoverable from the tracker, and a later session can't tell whether the ticket needs a human.

A session claims a ticket by assigning it to the dev driving the map, first, before any work, so concurrent sessions skip it. That assignee is the claim: an open, unassigned ticket is unclaimed.

Blocking uses the tracker's native dependency relationship — essential because it renders the frontier visually in the tracker's own UI, so the human sees what's takeable without opening the map. Only a tracker that lacks native blocking falls back to a body convention. A ticket is unblocked when every ticket blocking it is closed; the frontier is the open, unblocked, unclaimed children — the edge of the known.

The answer isn't part of the body — it's recorded on resolution (see Work through the map). Assets created while resolving a ticket are linked from the issue, not pasted in.

Ticket Types

Every ticket is either HITL — human in the loop, worked with a human who speaks for themselves — or AFK, driven by the agent alone. A HITL ticket only resolves through that live exchange; the agent never stands in for the human's side of it (a grilling agent that answers its own questions has broken this).

• Research (AFK): Reading documentation, third-party APIs, or local resources like knowledge bases to surface a fact a decision waits on. Resolved by the research skill — dispatched to a sub-agent where the harness has them, run inline where it doesn't. Use when knowledge outside the current working directory is required.
• Prototype (HITL): Raise the fidelity of the discussion by making a cheap, rough, concrete artifact to react to — an outline, a rough take, a stub, or UI/logic code via the prototype skill. Links the prototype as an asset. Use when "how should it look" or "how should it behave" is the key question.
• Grilling (HITL): Conversation. The default case. Always invoke the grill-me skill, and keep the domain language sharp as you go — the discipline ubiquitous-language applies: name a new concept in CONTEXT.md the moment it's settled, and challenge any term that conflicts with what's already there.
• Task (HITL or AFK): Manual work that must happen before a decision can be made — nothing to decide, prototype, or research, but the discussion is blocked until it's done. Signing up for a service so its API can be judged, provisioning access, moving data so its shape can be seen. This is the one type that does rather than decides — and it earns its place by unblocking a decision, not by delivering the destination. The agent drives it alone where it can (AFK); otherwise it hands the human a precise checklist (HITL). Resolved when the work is done; the answer records what was done and any resulting facts (credentials location, new URLs, row counts) later tickets depend on.

Fog of war

The map is deliberately incomplete: don't chart what you can't yet see. Beyond the live tickets lies the fog of war — the dim view of decisions and investigations you can tell are coming but can't yet pin down, because they hang on questions still open. Resolving a ticket clears the fog ahead of it, graduating whatever's now specifiable into fresh tickets — one at a time, until the way to the destination is clear and no tickets remain.

The map's Not yet specified section is where that dim view is written down: the suspected question, the area to revisit later. It's the undiscovered frontier toward the destination — everything here is in scope, just not sharp enough to ticket. Write as loosely or as fully as the view allows; it doubles as a signpost for collaborators reading where the effort is headed.

Fog or ticket? The test is whether you can state the question precisely now — not whether you can answer it now.

• Ticket when the question is already sharp — even if it's blocked and you can't act on it yet.
• Not yet specified when you can't yet phrase it that sharply. Don't pre-slice the fog into ticket-sized pieces: it's coarser than a ticket, and one patch may graduate into several tickets, or none, once the frontier reaches it.

Not yet specified excludes what's already decided (Decisions so far), what's already a live ticket, and what's out of scope (the next section).

Out of scope

Fog only ever gathers toward the destination. The destination fixes the scope, so work beyond it is out of scope — it isn't fog, and it doesn't belong in Not yet specified. It gets its own Out of scope section on the map: work you've consciously ruled out of this effort. Scope, not sharpness, lands it here.

Out-of-scope work never graduates — the frontier stops at the destination — so it returns only if the destination is redrawn, and then as a fresh effort, not a resumption.

Ruling something out of scope is a scoping act, not a step on the route. When a ticket that already exists turns out to sit past the destination — mis-scoped in while charting, or exposed by a resolution — close it (a closed ticket is unambiguously off the frontier) and leave one line in the Out of scope section: the gist plus why it's out of scope, linking the closed ticket. It stays out of Decisions so far, which records the route actually walked — a scope boundary isn't a step on it.

Invocation

Three modes. In Chart the map and Work through the map, never resolve more than one ticket per session — with the exception of AFK tickets, which spend no human attention. Ultra wayfinder resolves no HITL ticket at all; it dispatches and stops.

Offer the frontier

Charting and working both reach a moment where the frontier is workable. Stop and offer — never pick for the human.

List every frontier ticket by name, each with its type and whether it's HITL or AFK, and ask which to take. Their answer sets the batch: all of them, a named few, a count. Attention and resources are finite, that limit isn't visible from here, and so it is theirs to set rather than yours to guess.

Nothing runs until they answer — research included. Firing research unasked is cheap and usually the obvious call, which is exactly what makes it surprising the one time it isn't.

If the answer takes in several HITL tickets at once, that's Ultra wayfinder.

Chart the map

User invokes with a loose idea.

1. Name the destination. Run a grill-me session, sharpening the domain language as it goes, to pin down what this map is finding its way to — the spec, decision, or change. The destination fixes the scope, so it's settled first.
2. Map the frontier. Grill again, breadth-first this time: fan out across the whole space rather than deep on any one thread, surfacing the open decisions and the first steps takeable now. If this surfaces no fog — the way to the destination is already clear, the whole journey small enough for one session — you don't need a map. Stop and ask the user how they'd like to proceed.
3. Create the map (label wayfinder:map): Destination and Notes filled in, Decisions-so-far empty, the fog sketched into Not yet specified.
4. Create the tickets you can specify now as child issues of the map — then wire blocking edges in a second pass (issues need ids before they can reference each other). Wiring sorts them into the frontier and the blocked; everything you can't yet specify stays in the fog — the Not yet specified section.
5. Offer the frontier (above) rather than firing anything yourself. Act on the answer: AFK tickets run as sub-agents here — research ones via the research skill, capturing findings on a throwaway research/<name> branch with a context pointer from the ticket — and HITL tickets go out as peer sessions per Ultra wayfinder.
6. Stop — charting is one session's work; it hand-resolves no HITL ticket of its own.

Work through the map

User invokes with a map (URL or number). A ticket is optional — without one you work out what's next from the frontier rather than being handed it.

1. Load the map — the low-res view, not every ticket body.
2. Reconcile first. Any child ticket closed without a pointer in Decisions-so-far was resolved by a dispatched session, which is barred from writing the map (see Ultra wayfinder). Fold each one's resolution comment into the index, and act on any fog graduation or out-of-scope ruling it proposed. The map is the index; an unreconciled map has stopped indexing.
3. Choose the ticket. If the user named one, use it. Otherwise compute the frontier: one workable ticket, take it; more than one, offer the frontier (above). If they choose several HITL tickets, that's Ultra wayfinder and it claims per session — go there instead of continuing here. Otherwise claim the single ticket: assign it to yourself before any work.
4. Resolve it — zoom as needed: fetch the full body of any related or closed ticket on demand; invoke the skills the ## Notes block names. If in doubt, use grill-me.
5. Record the resolution: post the answer as a resolution comment, close the issue, and append a context pointer to the map's Decisions-so-far. Dispatched sessions stop at the comment — see Ultra wayfinder.
6. Add newly-surfaced tickets (create-then-wire); graduate any fog the answer has made specifiable, clearing each graduated patch from Not yet specified so it lives only as its new ticket. If the answer reveals a ticket — this one or another — sits beyond the destination, rule it out of scope rather than resolving it on the route. If the decision invalidates other parts of the map, update or delete those tickets. A dispatched session proposes all of this in its resolution comment instead.
7. If clearing that fog widened the frontier, offer it (above) before the session ends.

The user may run unblocked tickets in parallel, so expect other sessions to be editing the tracker concurrently.

Once the way is clear, the map is done. Tell the user to run to-tickets next — it cuts the actual build into tracer-bullet slices, and it is theirs to invoke, not yours.

Ultra wayfinder — work the frontier in parallel

Capability required: a harness that can dispatch peer sessions — sibling agent sessions the human attaches to and converses with directly. Sub-agents are not a substitute, for the reason below. Where the harness has none, Ultra is unavailable: say so plainly and work the frontier serially.

The user invokes it with "ultra wayfinder" (or "ultrawayfinder") plus a map, and optionally the tickets to take. Named tickets are the batch. Unnamed, it still offers the frontier first — invoking Ultra says how the work should run, not that you may pick how much of it. Reaching Ultra through Offer the frontier means the batch is already whatever the human chose there.

Peer sessions, not sub-agents. A sub-agent cannot do a HITL ticket: grilling and prototype tickets resolve only through live exchange with the human, and a sub-agent has no way to reach them — so it ends up answering its own questions and posting a resolution with none of the human's judgement in it. A dispatched peer session parks awaiting input and the human attaches to each in turn. That buys parallel state — every ticket's context loaded and warm — against serial human attention, which is the scarce thing.

Dispatch on the HITL/AFK axis, never on the type label — the axis is what decides whether a session can stand in for a human at all:

• HITL (grilling, prototype, task marked HITL) → one peer session each.
• AFK (research, task the agent drives alone) → a sub-agent in this session. There's no human to attach to, so a peer session would buy nothing and cost a context switch.

1. Load the map and compute the frontier.
2. Take the human's chosen set. Never fan out wider than they asked for.
3. Dispatch each HITL ticket to a peer session, carrying the brief below.
4. Run each AFK ticket as a sub-agent here, and record its resolution on the map yourself.
5. Confirm every dispatched session claimed its ticket. A session that died before its first write leaves the ticket unclaimed and back on the frontier while the human has been told it's in flight — claiming is the only concurrency guard, so it's the claim you verify, not the session. Re-dispatch or report any that didn't land.
6. Report the dispatched set by name, then stop.

The brief

Every dispatched session gets these four things, and the first is what makes the rest safe:

1. You are a dispatched session. Say it in those words. The session lands in Work through the map and will otherwise follow step 5 verbatim and write the map — which is exactly what must not happen.
2. Invoke wayfinder explicitly, by its command form. This skill sets disable-model-invocation, so a session merely told to use wayfinder loads none of it — including claim-first. The failure is invisible: the session starts, parks, and looks correct. (epic-runner handles its own dispatch the same way.)
3. Claim the ticket first, before any work. If it is already assigned, someone got there first — stop and report back rather than working it.
4. The map and the ticket, and: work up to the first question that genuinely needs the human, then stop and wait.

Dispatched sessions must not edit the map body. They post their resolution comment and close their issue as normal, but appending to Decisions-so-far is a read-modify-write on one shared document with no locking, and concurrent sessions silently clobber each other. So they propose — resolutions, fog graduations, out-of-scope rulings all land in the resolution comment, and the next session to work the map folds them in at its reconcile step.

AFK sub-agents are the exception, and it's the locking that makes them one: they run inside the dispatching session, so there is exactly one writer and you record their resolutions normally.

Example binding — Claude Code:

claude --bg -n "wf-<number>-<short-slug>" "/wayfinder <map> <ticket> — you are a dispatched session; <rest of brief>"

Pass -n always: without it the session's row is titled with the raw prompt text and the list becomes unreadable. Verify with claude agents --json. Dispatched sessions park in "Needs input" in Agent View (← on an empty prompt, or claude agents), and the human attaches with →.

Adapted from mattpocock/skills, MIT © 2026 Matt Pocock. Keep this line if you copy this file.