Content
67%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-structured, genuinely actionable instruction skill: concrete templates, exact labels, and clearly sequenced two-mode workflows with explicit checkpoints. Its costs are stylistic verbosity from the wayfinding metaphor and two unresolvable pointers to tracker documentation that is not part of the bundle.
Suggestions
Trim the metaphorical narration (e.g. "wrapped in fog", "the dim view of decisions and investigations you can tell are coming") to plain statements, and state each concept once instead of defining Not yet specified and Out of scope both in the map-body template and in their own sections.
Resolve the dangling pointers: either bundle a reference for the local-markdown tracker and the tracker "Wayfinding operations" conventions, or inline the minimal operations needed when no tracker doc is provided.
Add a concurrency checkpoint to the work-through-the-map steps (e.g. re-check the ticket is still open and unclaimed immediately before resolving), since the body itself warns that other sessions edit the tracker concurrently.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body teaches genuinely novel concepts (fog of war, frontier, claim protocol, HITL/AFK) without explaining anything Claude already knows, but the metaphor-heavy narration ("wrapped in fog", "the dim view of decisions", "the edge of the known") and repeated definitions (Not yet specified and Out of scope are each defined both in the map-body template and again in their own sections) add tokens that could be trimmed. It is above anchor 2 because the padding is stylistic rather than redundant explanation of known concepts. | 3 / 5 |
Actionability | Concrete, executable guidance throughout: full markdown templates for the map body and ticket body, exact labels (`wayfinder:map`, `wayfinder:<type>`), named Skill-tool invocations ("grilling", "domain-modeling", "research", "prototype"), and numbered step sequences for both invocation modes. It falls short of 5 because two pointers dangle: the "tracker doc's 'Wayfinding operations' section" and the "local-markdown tracker" fallback reference resources not present in this skill's bundle, leaving a session without the concrete operations they specify. | 4 / 5 |
Workflow Clarity | Both invocation modes are cleanly numbered with explicit sequencing checkpoints: claim before any work, create-then-wire in a second pass, never resolve more than one ticket per session, and post-resolution-close-append as a fixed record sequence. It is below 5 because there is no checkpoint for re-verifying a ticket's state before acting on a tracker that other sessions may be editing concurrently, though the one-ticket-per-session rule partially covers this; it is not a destructive or batch operation, so the 3-cap does not apply. | 4 / 5 |
Progressive Disclosure | A single-file instruction skill with no bundle files; every reference in the body is an internal anchor ([Ticket Types](#ticket-types), [Work through the map](#work-through-the-map)) and the templates sit inline where they belong, with sections well-organized for navigation. It sits just under 5 because the body is ~128 dense lines whose Ticket Types detail could live in a reference file, keeping the top-level map of the skill leaner. | 4 / 5 |
Total | 15 / 20 Passed |