Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/firedmosquito831-export
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/export
What's inside
- Skill name
export- Description
- Load for /vellum:export. The export skill is the operator's manual for building the final manuscript: it runs the readiness gate, interprets a failure as a prioritized fix list, runs the deterministic build on PASS, and explains the manifest and the pandoc fallback. The engine implementation lives in scripts/vellum_lib/export.py (Workstream B); this skill is what the operator (muse) follows.
- License
- Apache-2.0
- Instructions
- 44 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- files: referenced but missing: scripts/vellum_lib/export.py
Files (3)
- SKILL.md
- assembly.md · resources/
- manifest.md · resources/
SKILL.md (first lines)
# Export The export skill is the operator's manual for `/vellum:export` — the command that builds the final manuscript artifact. It is **not** the build engine (that is `scripts/vellum_lib/export.py`, Workstream B); it is the sequence the operator follows, the gate it runs first, and the way it interprets success and failure. The manuscript on disk is untouched — export is a build artifact, assembled from the gated sources. ## The export sequence 1. **Run `vellum readiness`.** This evaluates the preconditions (`gates/resources/readiness.md`): every chapter `final` with complete frontmatter, ledgers clean, the readiness report PASSES, the word-budget report attached, pivotal chapters have blind artifacts. 2. **If PASS** → run `vellum export build --out export/manuscript-<date> [--epub]`. Interpret the manifest (`resources/manifest.md`). 3. **If not PASS** → readiness prints a prioritized missing list. Present it to the author; do not run the build. Muse does not re-litigate the gate. Muse runs `vellum readiness`, reads its output, and either runs the build or presents the fix list. The gate is the engine's verdict. ## Interpreting a failure A failed readiness check prints a prioritized fix list. The format mirrors the beta-reader REVISE output: the changes that would most raise the verdict, each with location, evidence, intervention kind, and the axis/gate it moves. The author works the list, re-runs `/vellum:export`, and the gate re-evaluates. Common failures: - **Incomplete frontmatter** on a chapter → `vellum ledger check` emits `frontmatter:incomplete`; the author (or `@kb-lead`) fills the fields. - **Open ledger findings** → resolve or dismiss-with-author-signoff (`vellum dismiss`). - **Readiness report REVISE** → the beta reader's prioritized fix list; revise and re-run the beta read. - **Pivotal chapter missing blind artifact** → run the blind read for that chapter. - **Cold-read BLOCKER** → resolve or triage. ## On PASS — the build `vellum export build --out export/manuscript-<date>` assembles (`resources/assembly.md`): - Title page, copyright placeholder, front/back matter, joined chapters (in frontmatter `number` order), and the manifest. - `--epub` adds a stdlib-zipfile EPUB with a stable `urn:uuid:<book-uuid>` identifier from `kb/story.md` so highlights survive rebuilds. **No pandoc required** — the EPUB writer is stdlib zipfile. - DOCX/PDF via pandoc when present; otherwise the markdown bundle + instructions. Pandoc is optional, never on the critical path. ## The manifest The build writes `manifest.md` (`resources/manifest.md`): chapter list, word counts, sha256 checksums, and **gate provenance** for every chapter (outline-approval date, blind verdict, beta scores, logged overrides). Export never silently drops a gate — the provenance is attached. The manifest is the build's certificate; it is not the manuscript. ## Load the resource needed - `resources/assembly.md` — the deterministic assembly phase: order of parts, chapter-joining conventions, scene-break rules, the output files, and gate provenance. - `resources/manifest.md` — the manifest fields: build metadata, chapter list, per-chapter gate provenance, book-level gates, overrides, and checksums. ## What export is not Export does not edit the manuscript, does not re-litigate the gates, and does not replace the readiness report or the blind/beta artifacts — it references them. The manuscript is the source of truth; the export is a build artifact assembled from it. If the manuscript changes after export, the author re-runs the build (the manifest's checksums detect the drift).
From FiredMosquito831/my-writing-companion, Apache-2.0.
More from FiredMosquito831/my-writing-companion
- 0.557score
qi-layer
FiredMosquito831
Use when writing or maintaining AGENTS.md, .context/CONTEXT.md, or CLAUDE.md mirrors: keep intent docs minimal and load-bearing.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
cold-read
FiredMosquito831
Load for /vellum:cold-read, after every major editorial program, and before export.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
demolition
FiredMosquito831
Load when the author invokes demolition on a chapter draft, or when muse offers it before marking a chapter final.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
gates
FiredMosquito831
Load when muse accepts a pivotal chapter, before export, or for any scoring task.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
kb-integrity
FiredMosquito831
Load when a vellum check failed, or when muse routes integrity work.
SkillBFreshness: B★ 2Claude Code +9 - 0.550score
project-setup
FiredMosquito831
One-time project setup for creative writing.
SkillBFreshness: B★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 9, 2026 (1 month ago)
- Last release
- Sep 11, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- Website
- vellum.danubelabs.net
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from FiredMosquito831
- 0.560score
config
FiredMosquito831
Configure prompt-cache warming (keep-alive pings for idle Claude Code sessions) per scope - global, CLI, desktop app, project folder or single session -…
SkillAFreshness: AClaude Code +9 - 0.557score
cold-read
FiredMosquito831
Load for /vellum:cold-read, after every major editorial program, and before export.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
gates
FiredMosquito831
Load when muse accepts a pivotal chapter, before export, or for any scoring task.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
demolition
FiredMosquito831
Load when the author invokes demolition on a chapter draft, or when muse offers it before marking a chapter final.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
qi-layer
FiredMosquito831
Use when writing or maintaining AGENTS.md, .context/CONTEXT.md, or CLAUDE.md mirrors: keep intent docs minimal and load-bearing.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
kb-integrity
FiredMosquito831
Load when a vellum check failed, or when muse routes integrity work.
SkillBFreshness: B★ 2Claude Code +9
Cite this page
export skill. Agent Mart. https://agentmart.dev/skills/firedmosquito831-export (last updated 2026-10-24)