agentleFS
pnppnp/sharepoint-skills/Skills/dossier/dossier/SKILL.md

dossier

Skill142 starsChanged 3 months agoMITView on GitHub

Renders a polished, self-contained HTML briefing from any data source — SharePoint lists, uploaded documents, or a verbal description. Use when asked to build a competitor landscape, account brief, candidate summary, site inventory, board bio, speaker profile, partner overview, or any grid of profile cards or single-subject one-pager. Interviews the user if data is incomplete, constructs a validated JSON document, then renders it into a sandbox-safe HTML file using grid mode (2–8 subjects as responsive cards) or one-pager mode (1 subject as a rich briefing). No external dependencies — output runs inside a SharePoint sandboxed iframe.

10 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 10 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check what the user has already provided before asking anything:

    line 23
  2. Very specific:

    SharePoint lists: Call list_items on the relevant lists.

    line 25
  3. Distinctive:

    If the data or intent is not fully clear, ask these questions in a single message — not one at a time:

    line 33
  4. Distinctive:

    Narrative emphasis: Should the summary emphasize product/capabilities, financials, relationships, risks, or something else?

    line 39
  5. Distinctive:

    Do not ask about mode unless the count is exactly 1 and the user has not indicated a preference.

    line 55
  6. Distinctive:

    Run every check before rendering.

    line 104
  7. Distinctive:

    No external URLs, CDN links, or web fonts in the output HTML.

    line 142
  8. Distinctive:

    All values must be pre-rendered — the page must be fully correct with JavaScript disabled.

    line 144
  9. Distinctive:

    Do not generate Python, PowerShell, or any other scripts — output is always the HTML file.

    line 145
  10. Distinctive:

    Do not render until the validation checklist passes.

    line 146
  11. Distinctive:

    Do not use SharePoint Person column types in any accompanying list schema — use text columns only.

    line 147

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “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

Dossier

Generates a complete, self-contained HTML briefing by (1) interviewing the user or parsing their data, (2) constructing and validating a JSON document, then (3) rendering that JSON into HTML using assets/dossier-template.html. The JSON is the authoritative source of truth; the HTML is fully determined by it.


Step 1: Determine the Data Source

Check what the user has already provided before asking anything:

  • SharePoint lists: Call list_items on the relevant lists. Read all needed columns before proceeding.
  • Uploaded documents or tables pasted in chat: Parse all content. Do not ask for data you already have.
  • Verbal description only: Proceed to Step 2.

Step 2: Interview the User

If the data or intent is not fully clear, ask these questions in a single message — not one at a time:

  1. Subject type: What are you profiling? (competitors, accounts/prospects, candidates, SharePoint sites, board members, speakers, partners, vendors — or something else)
  2. Subject count: How many subjects? This determines the rendering mode.
  3. Purpose and audience: What decision does this briefing support? Who reads it? This determines detail level and which key facts matter most.
  4. Key facts: What are the 3–6 structured facts most important for each subject? Examples for competitors: HQ, founded, ARR, headcount. For candidates: current role, tenure, location, skills. For sites: owner, content type, last activity.
  5. Narrative emphasis: Should the summary emphasize product/capabilities, financials, relationships, risks, or something else?
  6. Palette: Warm paper (cream/terra-cotta), deep ink (dark/gold), clean white (white/green), or slate (gray/blue)? Default to warm paper.
  7. Mode override (only ask if count is 1 or user seems to want detail): Do you want a full one-pager briefing or a compact card?

Confirm your understanding before building the JSON.


Step 3: Choose the Rendering Mode

Auto-detection rule:

  • 2–8 subjects → grid mode
  • 1 subject → one-pager mode
  • User explicitly says "one-pager" or "full briefing" → one-pager mode regardless of count
  • User explicitly says "grid" or "landscape" → grid mode

Do not ask about mode unless the count is exactly 1 and the user has not indicated a preference.


Step 4: Decide What Goes Where

Key facts hold structured, scannable, low-word-count data — the kind of facts a reader needs in 5 seconds. Rules:

  • 3–6 pairs. No sentences. Values should be short (under 6 words).
  • Good: "HQ | Austin, TX", "ARR | ~$40M est.", "Founded | 2019"
  • Bad: "Company description | This is a company that..." (belongs in summary)

Tags are the fastest signal — 2–5 short labels that classify the subject. Use for: industry vertical, stage, geography, product category, or any grouping the reader uses for orientation.

Summary is 2–5 sentences of prose. Write in third person. Lead with the most important fact. No bullet points. This is the editorial paragraph — what an analyst would write.

Activity holds the 2–5 most recent, most relevant developments — in most-recent-first order. Each item gets a short date prefix (e.g., "Apr 2026") and a one-sentence description. For competitors: product launches, fundraises, acquisitions, executive changes. For candidates: recent projects, promotions, publications. For sites: content updates, ownership changes, traffic milestones.


Step 5: Build and Validate the JSON Document

Construct the JSON object. Top-level structure:

report object (document metadata):

FieldRequiredDescription
titleyesPage title, e.g. "Competitor Landscape — Q2 2026"
subtitlenoOptional subtitle shown below title
paletteyeswarm-paper, deep-ink, clean-white, or slate
modeno"grid", "one-pager", or "auto" (default: "auto")
footernoFooter text, e.g. "Prepared May 28, 2026 · Confidential"

subjects array — ordered list of subject objects (1–8 items):

FieldRequiredDescription
idyesUnique slug, e.g. "acme-corp"
nameyesSubject display name
typeyesCategory label shown as a pill, e.g. "Competitor", "Prospect", "Enterprise Site"
statusnoStatus label text, e.g. "Active", "Watch", "Inactive"
status_colornogreen, amber, red, gray, or blue
key_factsyesArray of {label, value} pairs, 2–6 items
tagsyesArray of strings, 1–6 tags
summaryyes2–5 sentence prose paragraph
activitynoArray of {date, text} items, 0–5, most-recent first

Validation Checklist

Run every check before rendering. Fix any failure before proceeding.

Document level:

  • report.palette is one of: warm-paper, deep-ink, clean-white, slate
  • subjects array has 1–8 entries
  • report.mode is "grid", "one-pager", or "auto" (or omitted)

Per subject:

  • id is present and unique
  • name and type are present and non-empty
  • key_facts has 2–6 items; each item has label and value
  • tags has 1–6 items
  • summary is present and 2–5 sentences
  • status_color (if present) is one of: green, amber, red, gray, blue
  • activity items (if present) each have date and text; at most 5 items

Data integrity:

  • No key_facts value is a sentence — values must be short labels
  • Activity items are ordered most-recent first
  • Subject count matches the chosen mode (grid: 2–8, one-pager: 1; or mode is explicitly set)

Step 6: Render the HTML

  1. Read assets/dossier-template.html.
  2. Replace {{REPORT_TITLE}} in <title> with a concise page title.
  3. Replace {{PALETTE_CLASS}} on <body> with palette- + the palette name.
  4. Determine final mode: if report.mode is "auto" or omitted, use "grid" for 2–8 subjects and "one-pager" for 1 subject.
  5. Replace {{PAGE_HEADER}} with the .rpt-header strip (title + optional subtitle).
  6. Replace {{CONTENT}} with the rendered subjects — either the grid of cards or the one-pager layout. See references/components.md for exact HTML patterns.
  7. Replace {{FOOTER}} with the footer text (or empty string).
  8. Deliver the complete file as the response. No prose outside the file.

Hard Constraints

  • No external URLs, CDN links, or web fonts in the output HTML.
  • No <script src> tags. All JavaScript must be inline.
  • All values must be pre-rendered — the page must be fully correct with JavaScript disabled.
  • Do not generate Python, PowerShell, or any other scripts — output is always the HTML file.
  • Do not render until the validation checklist passes.
  • Do not use SharePoint Person column types in any accompanying list schema — use text columns only.

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.