agentsclimarketplace
Sign in

Bible assemble

Skill mytechsonamy/VibeFlow/skills/bible-assemble

Assembles the Product Bible into one coherent, navigable, shareable export at project end (Sprint 71) — the final "we shipped, here's the whole book" deliverable. Reads bible-status, builds a curated index (docs/product-bible/README.md) with a per-category table of contents linking every canonical doc (bible-native in place + VibeFlow-produced docs referenced where they live — never duplicated), assembles a single concatenated export (docs/product-bible/product-bible-export.md), runs a light cross-consistency curation pass (gaps, stale living docs, glossary↔domain-model↔PRD coherence), and writes .vibeflow/reports/bible-assembly.md with the completeness rollup + what's still owed. Read-mostly — writes only the bible index + export + report. Use at end of a cycle / project, or any time you want the current full bible.From its SKILL.md

Install
npx -y skills add mytechsonamy/VibeFlow --skill bible-assemble

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

4.1 KB, as published. Nobody here has run it

Bible Assemble — the final, curated Product Bible

Parts B (intake) and C (living-doc growth) fill and grow the bible. This is the assembly: turn the corpus into one coherent, navigable book you can hand to a new hire, a partner, or a sales team — and surface what's still missing or stale.

Step 1 — Read the corpus state

bash "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/bible-status.sh"

Gives present/missing per doc + per category. Assembly indexes what's present and flags what's owed.

Step 2 — Build the index (docs/product-bible/README.md)

Write a curated index with a per-category table of contents (business/GTM → product → engineering), and for each doc a row: title · lifecycle (living/static) · status (✓ present / ✗ owed) · a one-line summary · a link to the canonical file. Link source: bible docs at their docs/product-bible/... path and source: vibeflow docs where they actually live (docs/architecture.md, docs/adr/, design/design-spec.md, .vibeflow/reports/...) — the index points at the single source, it never copies. Lead with a completeness line (Bible: 21/25 present — business 6/10, product 5/6, engineering 10/9).

Step 3 — Assemble the single-file export (docs/product-bible/product-bible-export.md)

Concatenate the present docs into one navigable document, in category/phase order, each under a clear heading with its provenance (path + lifecycle). For a large source: vibeflow doc, include a linked summary + the link rather than inlining the whole thing (keep the export readable; note it's an excerpt). This is the "one file to read/share" artifact; the canonical sources stay authoritative.

Step 4 — Curation pass (light cross-consistency)

Flag — don't auto-fix — coherence gaps across the corpus:

  • Owed docs — present in the taxonomy but missing (esp. living docs that should exist by this phase).
  • Stale living docs — a living doc not grown in the phases its growsIn covers (cross-check against .vibeflow/reports/bible-update.md).
  • Cross-doc coherence — a Glossary term with no home in the Domain Model; a Persona never referenced by a PRD; an ADR decision not reflected in API Standards. Surface as findings { finding, why, impact, confidence }; the operator (or bible-update) resolves them.

Step 5 — Report

Write .vibeflow/reports/bible-assembly.md: the completeness rollup, the curation findings, and the two written artifacts (index + export). End with a breadcrumb: fill the owed docs via /vibeflow:bible-intake (inputs) or by running the phase that produces them, then re-assemble.

▶ Next: /vibeflow:product-bible (scaffold/seed any owed docs, then re-run /vibeflow:bible-assemble)

Read-mostly

bible-assemble writes only docs/product-bible/README.md, docs/product-bible/product-bible-export.md, and its report. It never edits a source doc (bible-native or vibeflow) — assembly is a view over the corpus, and curation only flags.

See also

  • docs/PRODUCT-BIBLE.md — the corpus model.
  • /vibeflow:bible-intake (ingest) · /vibeflow:bible-update (grow) · /vibeflow:product-bible (status + scaffold).

Keep looking

Skills are one crate of 627,801. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.