officecli

name	officecli
description	Create, analyze, proofread, and modify Office documents (.docx, .xlsx, .pptx) using the officecli CLI tool. Use when the user wants to create, inspect, check formatting, find issues, add charts, or modify Office documents.

Note: This file is kept as an implementation/reference document for maintainers. Runtime skill loading should prefer skills/office_documents/.

officecli

AI-friendly CLI for .docx, .xlsx, .pptx. Single binary, no dependencies, no Office installation needed.

Install

If officecli is not installed:

macOS / Linux

if ! command -v officecli >/dev/null 2>&1; then
    curl -fsSL https://raw.githubusercontent.com/iOfficeAI/OfficeCLI/main/install.sh | bash
fi

Windows (PowerShell)

if (-not (Get-Command officecli -ErrorAction SilentlyContinue)) {
    irm https://raw.githubusercontent.com/iOfficeAI/OfficeCLI/main/install.ps1 | iex
}

Verify: officecli --version

If officecli is still not found after first install, open a new terminal and run the verify command again.

---

Strategy

L1 (read) → L2 (DOM edit) → L3 (raw XML). Always prefer higher layers. Add --json for structured output.

---

Help System (IMPORTANT)

When unsure about property names, value formats, or command syntax, ALWAYS run help instead of guessing. One help query is faster than guess-fail-retry loops.

Three-layer navigation — start from the deepest level you know:

officecli pptx set              # All settable elements and their properties
officecli pptx set shape        # Shape properties in detail
officecli pptx set shape.fill   # Specific property format and examples

Replace pptx with docx or xlsx. Commands: view, get, query, set, add, raw.

---

Performance: Resident Mode

Every command auto-starts a resident on first access (60s idle timeout) — file-lock conflicts are automatically avoided. Explicit open/close is still recommended for longer sessions (12min idle):

officecli open report.docx       # explicitly keep in memory
officecli set report.docx ...    # no file I/O overhead
officecli close report.docx      # save and release

Opt out of auto-start: OFFICECLI_NO_AUTO_RESIDENT=1. Skipping close still works (resident exits on idle), but explicit close guarantees the file is flushed before the next command reads it.

---

Quick Start

PPT:

officecli create slides.pptx
officecli add slides.pptx / --type slide --prop title="Q4 Report" --prop background=1A1A2E
officecli add slides.pptx '/slide[1]' --type shape --prop text="Revenue grew 25%" --prop x=2cm --prop y=5cm --prop font=Arial --prop size=24 --prop color=FFFFFF

Word:

officecli create report.docx
officecli add report.docx /body --type paragraph --prop text="Executive Summary" --prop style=Heading1
officecli add report.docx /body --type paragraph --prop text="Revenue increased by 25% year-over-year."

Excel:

officecli create data.xlsx
officecli set data.xlsx /Sheet1/A1 --prop value="Name" --prop bold=true
officecli set data.xlsx /Sheet1/A2 --prop value="Alice"

---

L1: Create, Read & Inspect

officecli create <file>               # Create blank .docx/.xlsx/.pptx (type from extension)
officecli view <file> <mode>          # outline | stats | issues | text | annotated
officecli get <file> <path> --depth N # Get a node and its children [--json]
officecli query <file> <selector>     # CSS-like query
officecli validate <file>             # Validate against OpenXML schema

view modes

Mode	Description	Useful flags
outline	Document structure
stats	Statistics (pages, words, shapes)
issues	Formatting/content/structure problems	--type format|content|structure, --limit N
text	Plain text extraction	--start N --end N, --max-lines N
annotated	Text with formatting annotations
html	Static HTML snapshot (.docx/.xlsx/.pptx) — writes to stdout	--browser (open in default browser), --page N (docx), --start N --end N (pptx slide range)

view html vs watch — both render the same HTML (shared *.HtmlPreview.cs renderer). Use view html for one-shot snapshots (CI artifacts, archival, diffing, piping to files); use watch when you need live refresh or browser-side click-to-select. view html needs no server/port.

officecli view report.docx html > snapshot.html       # snapshot to file
officecli view report.docx html --browser             # open in default browser

get

Any XML path via element localName. Use --depth N to expand children. Add --json for structured output. Default text output is grep-friendly single-line per node: path (type) "text" key=val key=val ...

officecli get report.docx '/body/p[3]' --depth 2 --json
officecli get slides.pptx '/slide[1]' --depth 1          # list all shapes on slide 1
officecli get data.xlsx '/Sheet1/B2' --json

Run officecli docx get / officecli xlsx get / officecli pptx get for all available paths.

Stable ID Addressing

Elements with stable IDs return @attr=value paths instead of positional indices. Prefer these in multi-step workflows — positional indices shift on insert/delete, stable IDs do not.

/slide[1]/shape[@id=550950021]                    # PPT shape
/slide[1]/table[@id=1388430425]/tr[1]/tc[2]       # PPT table
/body/p[@paraId=1A2B3C4D]                         # Word paragraph
/comments/comment[@commentId=1]                    # Word comment

Use returned paths directly for subsequent set/remove. PPT also accepts @name= (e.g. shape[@name=Title 1]), with morph !! prefix awareness. Elements without stable IDs (slide, run, tr/tc, row) fall back to positional indices. Run officecli <format> get for the full list.

query

CSS-like selectors: [attr=value], [attr!=value], [attr~=text], [attr>=value], [attr<=value], :contains("text"), :empty, :has(formula), :no-alt.

officecli query report.docx 'paragraph[style=Normal] > run[font!=Arial]'
officecli query slides.pptx 'shape[fill=FF0000]'

validate

officecli validate report.docx    # Check for schema errors
officecli validate slides.pptx    # Must pass before delivery

For large documents, ALWAYS use --max-lines to limit output.

---

Watch & Interactive Selection

Live HTML preview that auto-refreshes on every file change. Browsers can click / shift-click / box-drag to select shapes; the CLI can read the current browser selection and act on it.

officecli watch <file> [--port N]      # Start preview server (default port 18080)
officecli unwatch <file>               # Stop the preview server

Open the printed http://localhost:N URL in a browser. Click any element to select; shift/cmd/ctrl+click to multi-select; drag from empty space to box-select (rubber-band). PPT/Word uses blue outline; Excel uses native-style green selection with crosshair and rectangular range selection. Excel extras: double-click a cell to edit inline (shows formula, commits on Enter/Tab); drag a chart to reposition it.

get <file> selected — read what the user clicked

officecli get <file> selected [--json]

Returns the DocumentNodes for whatever is currently selected in the watching browser(s). Empty result if nothing selected. Exit code != 0 if no watch is running for this file.

Workflow — agent acts on what the user visually selected:

# User clicks shapes in the browser, then asks "make these red"
PATHS=$(officecli get deck.pptx selected --json | jq -r '.data.Results[].path')
for p in $PATHS; do
  officecli set deck.pptx "$p" --prop fill=FF0000
done

Key properties

• Selection survives file edits. Paths use the stable @id= form (e.g. /slide[1]/shape[@id=10000]), so editing other shapes — or even the selected one — does not lose the selection.
• All connected browsers share one selection. Opening the watch URL in two tabs gives a shared cursor; clicking in one updates highlights in the other. Last-write-wins.
• Same-file single-watch. A given file can have only one watch process at a time; the second watch <file> errors.
• Group shapes select as a whole. Clicking any shape inside a <group> selects the group container, not the inner shape. The CLI sees /slide[1]/group[@id=N]. Drilling into individual children of a group is not supported in v1.
• PPT and top-level Word. Selection / mark works on .pptx shapes, pictures, tables, charts, connectors, groups, and on .docx top-level paragraphs (<p>/<h1-6>/<li>/.empty) and top-level <table>. Inherited layout/master decorations (footers, logos) and Word nested elements (table cells, run-level) are not addressable. Excel .xlsx does not emit data-path — mark/selection on xlsx will always resolve to stale=true. Excel support is a v2 candidate.

Marks — edit proposals waiting for review

Marks are edit proposals waiting for review. Use mark when you (or the user) want to see, evaluate, and approve changes BEFORE they hit the file. Marks live in the watch process only — nothing is written to disk until a separate set pipeline applies them.

Decision tree — pick one:

• User doesn't need to confirm? → set directly (straight to disk). Marks are overkill for one-shot changes.
• User wants to review before changes apply? → mark (propose → review → set → mark goes stale).
• Just leaving a permanent annotation in the file? → add --type comment (Word native, persists in file).

Four-step lifecycle:

1. Propose — agent scans and creates marks with find + tofix + note.
2. Review — human opens the watch URL, sees highlights, decides what to accept.
3. Apply — a pipeline reads get-marks --json and runs real set commands for accepted items.
4. Stale — after the underlying text changes, the mark's find no longer matches; stale=true signals "this proposal has been handled".

officecli mark <file> <path> [--prop find=...] [--prop color=...] [--prop note=...] [--prop tofix=...] [--prop regex=true] [--json]
officecli unmark <file> [--path <p> | --all] [--json]
officecli get-marks <file> [--json]

Prop	Meaning
find	Literal text to highlight (or regex when regex=true; raw form find='r"[abc]"' also accepted). 500ms match timeout.
color	CSS color from whitelist: hex, rgb(...), or one of 22 named colors. Invalid rejected.
note	Free-form reviewer comment.
tofix	Structured proposed replacement value (drives the apply pipeline).
regex	true to switch find to regex.

Path must be data-path format from watch HTML: Word /body/p[N] or /body/table[N]; PPT /slide[N]/shape[@id=ID] (preferred) or /slide[N]/shape[N]. Excel is not supported in v1 (marks always resolve stale=true). Native query paths like /body/p[@paraId=...] will NOT resolve.

Worked example:

officecli watch report.docx &
# Propose
officecli mark report.docx /body/p[3] --prop find="资钱" --prop tofix="资金" --prop color=red --prop note="术语错误"
# Review — human eyeballs highlights in browser, unmarks bad proposals
# Apply — read accepted marks, run real set commands
officecli get-marks report.docx --json \
  | jq -r '(.marks // []) | .[] | select(.tofix != null) | [.path, .find, .tofix] | @tsv' \
  | while IFS=$'\t' read -r path find tofix; do
      officecli set report.docx "$path" --prop "find=$find" --prop "replace=$tofix"
    done

All mark commands support --json. For >3 mutations, wrap the apply loop in batch or open/close for performance.

---

L2: DOM Operations

set — modify properties

officecli set <file> <path> --prop key=value [--prop ...]

Any XML attribute is settable via element path (found via get --depth N) — even attributes not currently present.

Without find=, set applies format to the entire element. To target specific text within a paragraph, use find= (see find section below).

Run officecli <format> set for all settable elements. Run officecli <format> set <element> for detail.

Value formats:

Type	Format	Examples
Colors	Hex, named, RGB, theme	FF0000, red, rgb(255,0,0), accent1..accent6
Spacing	Unit-qualified	12pt, 0.5cm, 1.5x, 150%
Dimensions	EMU or suffixed	914400, 2.54cm, 1in, 72pt, 96px

find — format or replace matched text

Use find= with set to target specific text for formatting or replacement. Works the same in Word and PPT — just swap paths. Format props are separate --prop flags — do NOT nest them.

# Format matched text (auto-splits runs)
officecli set doc.docx '/body/p[1]' --prop find=weather --prop bold=true --prop color=red

# Regex matching
officecli set doc.docx '/body/p[1]' --prop 'find=\d+%' --prop regex=true --prop color=red

# Replace text (use `/` for whole-document scope)
officecli set doc.docx / --prop find=draft --prop replace=final

# Replace + format
officecli set doc.docx '/body/p[1]' --prop find=TODO --prop replace=DONE --prop bold=true

# PPT — same syntax, different paths
officecli set slides.pptx / --prop find=draft --prop replace=final

Path controls search scope: / = whole document, /body/p[1] or /slide[N]/shape[M] = specific element, /header[1] = header, /footer[1] = footer.

Notes:

• Case-sensitive by default. Case-insensitive: --prop 'find=(?i)error' --prop regex=true
• Matches work across run boundaries
• No match = no error (silent success). --json includes "matched": N
• Batch JSON regex: {"props":{"find":"\\d+%","regex":"true","color":"FF0000"}}
• Excel: only find + replace supported (no find + format props)

add — add elements or clone

officecli add <file> <parent> --type <type> [--prop ...]
officecli add <file> <parent> --type <type> --after <path> [--prop ...]   # insert after anchor
officecli add <file> <parent> --type <type> --before <path> [--prop ...]  # insert before anchor
officecli add <file> <parent> --type <type> --index N [--prop ...]        # insert at position (0-based, legacy)
officecli add <file> <parent> --from <path>                               # clone existing element

--after, --before, --index are mutually exclusive. No position flag = append to end.

Element types (with aliases):

Format	Types
pptx	slide, shape (textbox), picture (image/img), chart, table, row (tr), connector (connection/line), group, video (audio/media), equation (formula/math), notes, paragraph (para), run, zoom (slidezoom), ole (oleobject/object/embed)
docx	paragraph (para), run, table, row (tr), cell (td), image (picture/img), header, footer, section, bookmark, comment, footnote, endnote, formfield (text/checkbox/dropdown), sdt (contentcontrol), chart, equation (formula/math), field, hyperlink, style, toc, watermark, break (pagebreak/columnbreak), ole (oleobject/object/embed). Document protection: set / --prop protection=forms|readOnly|comments|trackedChanges|none
xlsx	sheet, row, cell, chart, image (picture), comment, table (listobject), namedrange (definedname), pivottable (pivot), sparkline, validation (datavalidation), autofilter, shape, textbox, databar/colorscale/iconset/formulacf (conditional formatting), ole (oleobject/object/embed — no Remove yet), csv (tsv). value="=SUM(...)" auto-detects as formula. Formulas auto-evaluated on write (150+ functions including VLOOKUP, SUMIF, IF, DATE, PMT, etc.)

Pivot tables (xlsx)

officecli add data.xlsx /Sheet1 --type pivottable \
  --prop source="Sheet1!A1:E100" --prop rows=Region,Category \
  --prop cols=Year --prop values="Sales:sum,Qty:count" \
  --prop grandTotals=rows --prop subtotals=off --prop sort=asc

Key props: rows, cols, values (Field:func[:showDataAs]), filters, source, position, layout (compact/outline/tabular), repeatLabels (true/false — repeat outer row labels on every data row), blankRows (true/false — insert blank line after each group), aggregate, showDataAs (percent_of_total/row/col, running_total), grandTotals (both/rows/cols/none), subtotals (on/off), sort (asc/desc/locale/locale-desc). Aggregators: sum, count, average, max, min, product, stdDev, stdDevp, var, varp, countNums. Date columns auto-group. Multiple data fields and N×N row/col hierarchies supported. Run officecli xlsx set pivottable for full property list.

Document-level properties (all formats)

officecli set doc.docx / --prop docDefaults.font=Arial --prop docDefaults.fontSize=11pt
officecli set doc.docx / --prop protection=forms --prop evenAndOddHeaders=true
officecli set data.xlsx / --prop calc.mode=manual --prop calc.refMode=r1c1
officecli set slides.pptx / --prop defaultFont=Arial --prop show.loop=true --prop print.what=handouts

Run officecli <format> set / for all available document-level properties (docDefaults, docGrid, CJK spacing, calc, print, show, theme, extended).

Text-anchored insert (--after find:X / --before find:X):

The --after and --before flags accept a find: prefix to locate an insertion point by text match within a paragraph.

# Insert run after matched text (inline, within the same paragraph)
officecli add doc.docx '/body/p[1]' --type run --after find:weather --prop text=" (sunny)"

# Insert table after matched text (block — auto-splits the paragraph)
officecli add doc.docx '/body/p[1]' --type table --after "find:First sentence." --prop rows=2 --prop cols=2

# Insert before matched text
officecli add doc.docx '/body/p[1]' --type run --before find:weather --prop text="["

• Inline types (run, picture, hyperlink...) insert within the paragraph
• Block types (table, paragraph) auto-split the paragraph and insert between the two halves

PPT text-anchored insert — same as Word, but PPT only supports inline types (run); block-type insertion is not supported.

officecli add slides.pptx '/slide[1]/shape[1]' --type run --after find:weather --prop text=" (sunny)"

Clone: officecli add <file> / --from '/slide[1]' — copies with all cross-part relationships.

Run officecli <format> add for all addable types and their properties.

move, swap, remove

officecli move <file> <path> [--to <parent>] [--index N] [--after <path>] [--before <path>]
officecli swap <file> <path1> <path2>
officecli remove <file> '/body/p[4]'

When using --after or --before, --to can be omitted — the target container is inferred from the anchor path.

batch — multiple operations in one save cycle

Stops on first error by default. Use --force to continue past errors.

# Via stdin
echo '[
  {"command":"set","path":"/Sheet1/A1","props":{"value":"Name","bold":"true"}},
  {"command":"set","path":"/Sheet1/B1","props":{"value":"Score","bold":"true"}}
]' | officecli batch data.xlsx --json

# Via --commands (inline) or --input (file)
officecli batch data.xlsx --commands '[{"op":"set","path":"/Sheet1/A1","props":{"value":"Done"}}]' --json
officecli batch data.xlsx --input updates.json --force --json

Batch supports: add, set, get, query, remove, move, swap, view, raw, raw-set, validate. Fields: command (or op), path, parent, type, from, to, index, after, before, props, selector, mode, depth, part, xpath, action, xml.

---

L3: Raw XML

Use when L2 cannot express what you need. No xmlns declarations needed — prefixes auto-registered.

officecli raw <file> <part>                          # view raw XML
officecli raw-set <file> <part> --xpath "..." --action replace --xml '<w:p>...</w:p>'
officecli add-part <file> <parent>                   # create new document part (returns rId)

raw-set actions: append, prepend, insertbefore, insertafter, replace, remove, setattr.

Run officecli <format> raw for available parts per format.

---

Common Pitfalls

Pitfall	Correct Approach
--name "foo"	❌ Use --prop name="foo" — all attributes go through --prop
x=-3cm	❌ Negative coordinates not supported. Use x=0cm or x=36cm
PPT shape[1] for content	❌ shape[1] is typically the title placeholder. Use shape[2] or higher for content shapes
/shape[myname]	❌ Name indexing not supported. Use numeric index: /shape[3]
Guessing property names	❌ Run officecli <format> set <element> to see exact names
Modifying an open file	❌ Close the file in PowerPoint/WPS first
\n in shell strings	❌ Use \\n for newlines in --prop text="..."
officecli set f.pptx /slide[1]	❌ Shell glob expands brackets. Always single-quote paths: '/slide[1]'

---

Specialized Skills

For complex scenarios, load the dedicated skill from skills/<skill-name>/SKILL.md:

Skill	Scope
officecli-docx	Word documents — reports, letters, memos
officecli-academic-paper	Academic papers with TOC, equations, footnotes, bibliography
officecli-pptx	Presentations — general slide decks
officecli-pitch-deck	Investor/product/sales decks with charts and callouts
morph-ppt	Morph-animated cinematic presentations
officecli-xlsx	Excel — financial models, trackers, formulas
officecli-data-dashboard	CSV/tabular data → Excel dashboards with charts, sparklines

---

Notes

• Paths are 1-based (XPath convention): '/body/p[3]' = third paragraph
• --index is 0-based (array convention): --index 0 = first position
• After modifications, verify with validate and/or view issues
• When unsure, run officecli <format> <command> [element[.property]] instead of guessing