Embed the lark-slides reference docs (xml-schema-quick-ref.md and each shortcut's command guide) directly into the CLI help output so that agents can discover XML syntax and shortcut usage without reading separate markdown files.
- PrepareDomainHelp: append XML schema quick reference for slides domain
- PrepareShortcutHelp: embed shortcut-specific reference docs for +create, +xml-get, +screenshot, +media-upload, +replace-slide, +replace-pages, +history-list, +history-revert, +history-revert-status
- Add contract tests for all reference mappings and re-render idempotency
- +screenshot excludes the shared XML reference (user feedback)
Form questions can now carry a visible_rule (display condition) so a question shows only when earlier questions match the rule. The rule shares the exact same structure as the view filter, so extract that structure into a single shared reference (lark-base-filter-condition.md) that both view-set-filter and visible_rule point to.
- create/update shortcuts: document visible_rule in --questions help and transcribe the questions body (including visible_rule) into dry-run output
- document that form question updates use full overwrite semantics and must preserve existing fields via read-modify-write
- skill refs: add visible_rule sections to form-questions create/update, note it is only needed when the user asks for a display condition, and clarify that the shared tuple filter protocol does not apply to data-query filters
- tests: pin flag help, verbatim visible_rule passthrough on create/update/list, and add dry-run E2E coverage
Co-authored-by: yballul-bytedance <273011618+yballul-bytedance@users.noreply.github.com>
Co-authored-by: TRAE CLI <noreply@bytedance.com>
* feat(drive): add +permission-get-setting shortcut
Add a Drive shortcut for reading public permission settings across supported documents, files, folders, and wiki nodes. Resolve URLs into typed resources, preserve permission_public output for machine consumers, and document the shortcut in the permission-governance workflow.
Key features:
- Infer resource type and token from supported Drive URLs while requiring --type for bare tokens
- Query the Drive v2 public permission endpoint with typed validation and user or bot identity
- Support folder permission inspection without recursing into child resources
- Add unit, dry-run E2E, live workflow, output, and skill guidance coverage
* fix(drive): harden permission get setting contract
Harden +permission-get-setting after review findings so callers receive only the documented permission payload and folder support is verified against the live workflow. This prevents malformed responses from being presented as permission settings and keeps the command guidance aligned with the shortcut contract.
Key fixes:
- Reject responses without data.permission_public instead of projecting arbitrary payload fields
- Render complete permission settings in pretty output and mark --token required
- Exercise a created Drive folder in the live workflow and add the command reference
- Correct folder resolution guidance while retaining the shortcut's documented URL forms
* feat/drive-folder-permission-get
* feat(drive): add +member-list shortcut
Add a Drive shortcut for listing collaborators on documents, files, folders, and wiki nodes. Resolve supported resource URLs into typed permission requests, preserve raw API data for machine consumers, and keep invalid flag combinations on typed validation paths.
Key features:
- Infer resource type and token from supported Drive URLs while requiring --type for bare tokens
- Validate optional member fields and wiki-only permission type filters
- Provide pretty output, skill guidance, unit coverage, and dry-run/live E2E workflows
- Read dry-run assertions from the standard data.api success envelope
* feat/drive-member-list
Consolidate the user-search identity note into SKILL.md, and clarify bot
handling across attendee flows: bots are virtual identities with no
free/busy semantics, no meeting-room seat, and no room preference, so
they must be excluded from +suggestion, +room-find, and the scheduling
free/busy check. Note in create/update that bots remain valid attendees.
Reframe the identity section around event ownership: use `--as user`
for the logged-in user's own events and `--as bot` for events the bot
creates or participates in, with matching `+agenda` examples.
* feat(drive): add title+body joint search guidance and pagination rules for Top N results
- Add new blockquote explaining combined title+body search: use a single
--query with both keywords instead of splitting into two searches
- Add rule for Top N results: N is an output cap, not --page-size; scan
up to 3 pages filtering by title and summary_highlighted, read body
only for title-matched candidates, stop early at N confirmed results
- Add quick-reference table row for folder-scoped title+body search
- Update pagination strategy rule to cover the 3-page cap for joint
search in addition to the existing 5-page limit for other scenarios
* feat(drive): clarify Top N search output limit
* feat(drive): clarify search filters share one call
---------
Co-authored-by: yballul-bytedance <273011618+yballul-bytedance@users.noreply.github.com>
Replace the uniform 0.55em half-width coefficient with per-character-type
coefficients, add font-family awareness (sans/serif), bold multiplier,
letter-spacing support, and fix padding-aware line wrapping.
- Split half-width chars into uppercase (0.57), lowercase (0.51 sans / 0.53
serif), digits (0.58), and punctuation (0.50)
- Add classify_font_family() to apply slightly wider lowercase widths for
serif fonts (Georgia, Source Han Serif/思源宋体, Times, etc.)
- Add 5% width multiplier for bold text; detect <strong>/<b>/<i>/<em> tags
and span-level bold/italic attributes in addition to content attrs
- Fix estimate_text_line_count_for_text to subtract paddingLeft/paddingRight
from available width before computing wrap lines
- Add resolve_letter_spacing and wire letterSpacing through estimate_text_width
- Extract fontFamily/bold/italic/letterSpacing into element dict during parse
chartParsedValues is a server-injected roundtrip child tag under
chartField, not an attribute. Move it from ROUNDTRIP_SXSD_ATTRS to a
new ROUNDTRIP_SXSD_TAGS set and skip the tag (and its subtree) in the
SXSD tag whitelist check.
Large low-alpha text underneath other text shapes is typically a
background design element; treat text_may_overflow_shape as info in
that case instead of warning/error.
Extract letterSpacing from content/paragraph attrs and factor it into
width and line-count estimates, and stop short-circuiting the shape
overflow check for autoFit shapes so that letterSpacing-heavy captions
under normal-auto-fit no longer escape detection.
Text-shape overflow was always reported as a warning, which let clearly
broken pages pass the lint gate. Overflow > 10px now upgrades to error;
smaller overflows stay as warning to avoid flagging near-fit cases.
--slide-id used the cobra StringArray flag type, which only accepts
repeated flags and does not split comma-separated values, unlike
--slide-number (int_array -> cobra IntSlice) which already supported
CSV input. This made the two selector flags inconsistent.
Switch --slide-id to the string_slice flag type (cobra StringSlice),
which natively supports both comma-separated and repeated values, and
update the flag readers from StrArray to StrSlice. normalizeSlideIDs
already trims/dedupes/filters blanks, and
validateSlidesScreenshotSelectorLimit already caps the combined
selector count, so both continue to apply unchanged to CSV input.
Add tests covering --slide-id CSV parsing, whitespace/duplicate
normalization, and the >10 selector limit via CSV, mirroring the
existing --slide-number coverage.
Address review feedback:
- Fix "comma-separate" -> "comma-separated" wording in the --slide-id
flag description (CodeRabbit).
- Set LARKSUITE_CLI_CONFIG_DIR to t.TempDir() in the new screenshot
tests, per the AGENTS.md testing convention, so local configuration
state cannot leak into or be modified by the suite.
- Add a dry-run E2E test (tests/cli_e2e/slides) that pins --slide-id
CSV parsing through the built CLI binary and asserts the emitted
slide_ids request body, per the AGENTS.md dry-run E2E requirement
for shortcut flag/param changes.
- Update the lark-slides skill reference to document that --slide-id
and --slide-number both accept comma-separated values, not just
repeated flags, so agents can discover the new syntax.
Form submission writes and submits data through a public share link, an
irreversible action that should require explicit confirmation. Reclassify
the shortcut from write to high-risk-write so the runner's --yes gate fires
before execution, matching +form-delete and other high-risk base commands.
Update the lark-base skill docs (--yes on all examples, param table, tips)
and add tests pinning the confirmation gate (unit) and dry-run structure (e2e).
Co-authored-by: yballul-bytedance <273011618+yballul-bytedance@users.noreply.github.com>
feat(slides): add layout density lint for sparse/empty containers
Extend the XML layout lint into a single release gate for Slides XML:
- Add blank_slide, sparse_container_content, and sparse_slide_content
detection, using visibility- and coverage-aware heuristics (alpha
filtering, image-overlay/layout-panel exemptions, similar-short-card
grouping) to avoid flagging intentional whitespace or background
panels
- Broaden out-of-canvas detection from table/chart/text-only to every
element kind, with rotation-aware bounding boxes and geometry
extraction for icon/line/polyline
- Restructure output to schema v2.0: every issue carries rule
(id/name/comparison/threshold), measurement, related_objects, and
hint; summary gains status/release_ready/screenshot_review_required
- Change CLI exit-code semantics so only errors block (exit 1);
warning-only output still exits 0 to let downstream screenshot review
proceed
- Harden XML attribute parsing (single/double-quoted and spaced
attributes, self-closing tags no longer bleeding content into the
next element) and fix edge cases surfaced during review
(image-overlay coverage ratio, invisible container/panel exemptions,
bbox_overlap measurement consistency, background-only slide bypass,
invisible short-card peers)
- Update SKILL.md, validation-checklist.md, and troubleshooting.md to
match the new gate; add regression tests for the new rules and fixes
* docs(base): clarify complete and partial updates
Consolidate the update rule introduced in #1879 and make the command-contract boundary explicit. Full-update commands must use trusted current configuration for the first actual request, while delta commands should send the smallest legal payload.
* docs(base): clarify full-update state preservation
Address review feedback by requiring unchanged writable configuration to remain intact, except when the requested update makes a setting inapplicable.
* docs(base): strengthen update contract guidance
From EVAL-07-22-02-53 (42 convos), agents fell back to the full XSD for:
- shape type enum + presetHandlers (rounded corners)
- polyline (bounding-box positioning, required border, connector type)
- table merged cells (colspan / rowspan)
Add compact coverage for each, sized to real usage (shape/polyline type
lists trimmed to what actually appears in generations). Chart gaps deferred.
paas_storage AppFileListForOpenAPI rejects page_size > 200 at the inner
checkMaxKeys guard with ErrInvalidRequest("maxKeys not in range (0, 200]").
Previously the CLI forwarded any --page-size straight to the API, so
--page-size 500 produced an opaque server error round-trip.
Add a client-side Validate check bounding --page-size to [1, 200] (aligned
with the existing validateAppsPageSize precedent in the observability
commands): out-of-range values now fail fast with a typed validation error
and never hit the network. The server tolerates page_size <= 0 by defaulting
to 20, but the CLI default is already 20 and an explicit < 1 is a user error,
so we reject it for a clearer message, consistent with other list commands.
Update the flag description and the lark-apps-file skill reference to
document the 1..200 range, and cover the boundaries in unit tests.
Card header icon documentation contained invalid tokens (e.g., mail_colorful, approve_colorful) that do not render, and icon guidance lacked precise token enumeration, causing LLM to guess or fabricate icon tokens. This PR replaces
examples with valid tokens and adds a definitive colorful icon reference table.
zhanghuanxu