agentleFS
indranilbanerjeeindranilbanerjee/socialforge/skills/assemble-document/SKILL.md

assemble-document

Skill43 starsChanged 4 months agoMITView on GitHub

Assemble the month-end delivery manifest: approved posts, copy, files and metadata as one JSON. \"build the delivery manifest\"

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create the final delivery manifest — a structured JSON file describing the complete monthly calendar, its posts, copy, and metadata.

    line 14
  2. Very specific:

    Read ai_disclosure from brand-config.json (missing block = {"mode": "claude-surfaces", "text": null}) and decide whether the manifest …

    line 34
  3. Very specific:

    Run python "${CLAUDE_PLUGIN_ROOT}/scripts/detect_surface.py" --mode {mode} — its disclosure_applies field IS the decision.

    line 36
  4. Very specific:

    When it applies, add to the manifest metadata: "ai_assistance_note": "Creative produced with AI assistance under human review and brand …

    line 37

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/socialforge:assemble-document — Document Assembler

Script location. If your host does not set ${CLAUDE_PLUGIN_ROOT}, the scripts are in this plugin's scripts/ folder, next to skills/. Run them with Python 3.10 or newer: python, or python3 where python is not found (macOS, many Linux systems).

Create the final delivery manifest — a structured JSON file describing the complete monthly calendar, its posts, copy, and metadata. The manifest is the deliverable this skill produces; a formatted DOCX is a manual/optional step downstream (see below).

Manifest Structure

assemble_docx.js emits these sections into the JSON manifest:

  1. Title block (brand name, month, generation timestamp)
  2. Monthly overview (post count, platform breakdown, tier distribution, content-type distribution)
  3. Weekly sections:
    • For each post: post id, date, title, tier, platforms, content type, copy option A, visual direction, creative mode, status
  4. Publishing schedule (date, day of week, post id, title, platforms)

Process

  1. Load calendar-data.json and status-tracker.json for the brand + month
  2. Group posts by week and merge in each post's tracked status and creative mode
  3. Build the manifest structure
  4. Save to ${CLAUDE_PLUGIN_DATA}/socialforge/output/{brand}/{month}/FINAL/00-Calendar-Document/{brand}-{month}-calendar.json (falls back to ~/socialforge-workspace/output/... when ${CLAUDE_PLUGIN_DATA} is unset)

AI-assistance note (delivery manifest)

Read ai_disclosure from brand-config.json (missing block = {"mode": "claude-surfaces", "text": null}) and decide whether the manifest carries the AI-assistance note:

  1. Run python "${CLAUDE_PLUGIN_ROOT}/scripts/detect_surface.py" --mode {mode} — its disclosure_applies field IS the decision. Fail-safe: an uncertain surface applies the note in claude-surfaces mode; skipping requires an AFFIRMATIVE non-Claude fingerprint. Never override the script's answer.
  2. When it applies, add to the manifest metadata: "ai_assistance_note": "Creative produced with AI assistance under human review and brand approval gates." (or the brand's custom text verbatim). The default wording is vendor-neutral and claims only the review this pipeline actually performs — every post passed the approval chain.
  3. Record the decision either way: "disclosure": {"applied": true|false, "mode": ..., "surface": ...} in the manifest — an unapplied note is a recorded choice, not an omission.
  4. Remind the user at handoff: platform-native AI-content labels (Instagram/TikTok/YouTube toggles) are the right place for per-post disclosure — flag which posts used AI generation so whoever publishes can set them.

Producing a DOCX (manual/optional)

No DOCX is generated automatically — the script reports docx package not available and emits JSON only. To produce a Word document, hand the manifest JSON to a document tool of your choice, or lay it out against assets/document-template/. Image previews are referenced by path in the manifest rather than embedded.

Timeout & Fallback

  • Manifest assembly: 2-minute timeout for 30 posts.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.