JoshuaShepherd Generate pathway page content: $ARGUMENTS
$ARGUMENTS should name the author and the pathway theme (and may add any standard
parameter below). Example: author: brad-brisco, theme: Bivocational Mission, reframing: "What if your job is not what you do to fund the mission, but part of the mission itself?". If the
author or theme is missing, ask for them before proceeding — they are the two required inputs.
What this skill does, and where it sits
A Movemental pathway page is the canonical twelve content sections authored in the tenant author's voice and grounded in that author's corpus. The page template (hub and individual page) already exists. This skill produces the thing the template needs: the markdown content of the twelve sections, written in the right author's voice, drawn from that author's books, with citations, the four pedagogical necessities discharged, and the standard parameters set.
It is the manual's "Phase 1 — Pathway design", generalized across authors and made
voice-aware and corpus-grounding. It runs before the code-level pathway-author skill (which
turns this markdown into src/lib/content/pathways/[slug].ts) and before pathway-audit.
It does not: build the HTML/React page template, write the code-level [slug].ts, configure
the AI Lab agent, or invent the author's positions. Stop at the reviewed markdown unless asked
to hand off to pathway-author.
The prime directive: the author's real voice, the author's real corpus
Two non-negotiables, inherited from the platform's publication gates:
- Corpus-grounding. Every framework, claim, story, historical example, and practice traces to something the author actually wrote or said, found in the repo. Never invent a position the author has not taken. If the corpus does not support a section, say so and narrow the section — do not fabricate.
- The author's own voice — not a generic "thought-leader" voice, and not another author's voice. Load this author's documented voice from the repo and apply it. Alan Hirsch does not sound like Brad Brisco; a pathway written in the wrong voice is a failure even if every fact is right.
If you cannot find the author's voice documentation or corpus in the repo, stop and tell the user where you looked — do not proceed on a guessed voice.
Standard parameters
Resolve these before generating. Required inputs are the first two; the rest are derived from the repo (and confirmed with the user) when not supplied.
| Parameter | Required | How to resolve if absent |
|---|---|---|
author |
yes | — (ask). A tenant/leader slug, e.g. alan-hirsch, brad-brisco |
theme / title |
yes | — (ask). The pathway's subject |
reframing_question |
recommended | Draft one from the theme + the author's framing of it in the corpus, then confirm with the user. This is the disturbance that opens the formation work |
slug |
derived | Kebab-case the theme (e.g. bivocational-mission) |
companion_pillar |
derived | Look up the author's pillar map (research / dashboard-content); ask if unclear |
primary_corpus |
derived | The book(s) the theme draws from; pick from the author's bibliography by theme match; confirm |
companion_course |
optional | Link if a matching course exists; otherwise omit |
output_path |
derived | Default content/pathways/<author>/<slug>.md (or the user's chosen location) |
Workflow
Work through these in order. Steps 1–2 are research; step 3 is generation; steps 4–5 are checking and output. Read the reference file named in each step before doing that step.
1. Resolve the author and load their voice
Read references/voice-application.md. Locate this author's voice documentation in the repo
(in priority order: a dedicated voice skill such as alan-voice if one exists; the author's
dashboard-content/ Part III — Voice & Editorial Identity; the raw
movement_leader_research/<author>/ voice files). Build a short working voice profile: the
author's markers, signature moves, anti-patterns, and the rhetorical posture. You will apply this
in step 3 and check against it in step 4.
2. Gather the corpus material for this theme
Read references/corpus-gathering.md. Identify the primary corpus draws (books) for the
theme, then gather, per canonical section, the specific material that section needs — the
framework's source chapter, the author's biblical engagement, the historical parallels, the
narrative cases, the embodied practices, the distortions the author names. Record each with a
citation (book, chapter, page where available). This becomes the citations block. Gather before
you write; do not write a section you have no corpus for.
3. Generate the twelve sections, in canonical order, applying voice where it matters
Read references/pathway-content-model.md — it specifies what each of the twelve sections
must contain, the four-necessities pedagogy, and the per-section voice intensity. Write the
sections in order into the template (assets/pathway-content-template.md). Apply the author's
voice in proportion to the section: the hero, model, scripture thread, historical context,
cases, practices, distortions, and invitation are voice-forward; the curated resources, FAQs,
and AI Lab invitation are lighter and more structural. Ground every substantive section in the
corpus with citations.
4. Self-check before output
Verify, in this order (details in references/pathway-content-model.md):
- Completeness — all twelve sections present and in order; cases are plural; distortion warnings present; AI Lab invitation present; invitation names one concrete next step.
- Pedagogy — all four necessities discharged: dissonance (hero/model/distortions), action (practices), reflection (structured reflection questions, not a journaling gesture), community (a named cohort or practice partnership in the invitation).
- Corpus-grounding — every substantive section cites ≥3 corpus sources; no invented position; every citation resolves in the citations block.
- Voice — the working voice profile's markers are present and its anti-patterns absent; it reads as this author, not a generic voice.
If any check fails, revise before delivering.
5. Output and hand off
Write the completed markdown to output_path. Tell the user what you produced and offer the
handoff: this markdown is the input to the code-level pathway-author skill (which writes
src/lib/content/pathways/<slug>.ts) and to pathway-audit. Do not run those unless asked.
Voice intensity by section (apply voice where relevant)
| Section | Voice intensity | Why |
|---|---|---|
| 1 Hero / provocation | Highest | The reframing question is the author's sharpest move |
| 2 Overview | High | Sets the frame in the author's terms |
| 3 Model / framework | High | The author's own framework, in their language |
| 4 Scripture thread | High | The author's biblical engagement, not generic exegesis |
| 5 Historical context | High | The author's chosen parallels and how they tell them |
| 6 Cases | High | Narrative is voice-bearing; tell the cases as the author tells them |
| 7 Practices | Highest | Embodied invitation in the author's pastoral/prophetic register |
| 8 Curated resources | Light | Mostly pointers and links; voice in the framing line only |
| 9 AI Lab invitation | Light–Medium | Short, in-voice invitation to converse |
| 10 FAQs | Medium | Answers in voice; questions plain |
| 11 Distortion warnings | High | The author's own naming of cheap/reduced versions |
| 12 Invitation | Highest | The closing call; voice and concreteness both peak |
Output format
Deliver one markdown file following assets/pathway-content-template.md: a parameter front-matter
block, the twelve sections in canonical order with their headings, inline citation markers, and a
citations block at the end that mirrors the footnotes-registry schema (claim, source, page, type).
State the author and theme used, the primary corpus drawn from, and note any section you had to
narrow for lack of corpus support.
Examples
Example 1 — full generate:
Input: author: brad-brisco, theme: Bivocational Mission
Action: resolve voice + corpus → draft a reframing question and confirm → gather from Brad's
books on work/vocation/mission → write twelve sections in Brad's voice → self-check → output
content/pathways/brad-brisco/bivocational-mission.md.
Example 2 — author named, theme loose:
Input: "do the APEST pathway for Alan"
Action: theme = APEST vocation; load Alan's voice (via alan-voice if present); primary corpus =
5Q and The Permanent Revolution; companion pillar = the APEST pillar; generate and output.
Example 3 — missing inputs: Input: "write a pathway page" Action: ask which author and which theme before doing anything else — they are the two required parameters.