Add explicit reviewed Agent hints for high-frequency sheet range/filter/filter-view, condition-format and dropdown tools. Replace generic avoid_when with concrete read/write/clear/style/filter-view disambiguation, tighten destructive operations, and regenerate schema metadata/catalog.
Validated with drift/catalog gates and go test ./internal/cli ./internal/app ./internal/generator/... .
Add explicit attendance Agent review hints for all 38 attendance tools, replacing template avoid_when with business-specific selection guidance and marking them reviewed. Tighten high-impact attendance writes such as boss-check and settings/balance updates with high risk and user confirmation.
Regenerate schema metadata/catalog and update parameter binding hash. Drift/catalog gates and key schema tests pass.
Stop registering runtime catalog fallback commands and make command-surface generation use the real Cobra tree directly. Regenerate schema surface, agent metadata and catalog from executable commands (20 products / 537 tools), add runtime-surface completeness hints, and update catalog gates/tests to use dynamic counts instead of old 504/21/461 constants.
This makes schema describe the actual executable CLI surface; drift/catalog gates and go test ./internal/cli ./internal/app ./internal/generator/... pass.
Add explicit reviewed summaries for aitable view get/update subcommands so Agents
can distinguish filter, sort, group, visible-fields, aggregate, card and other
view operations. Regenerate sibling-disambiguation avoid_when entries from the
new summaries, making cross-tool guidance precise instead of generic.
Results: 395/504 tools carry sibling-command disambiguation and reviewed coverage
rises to 104/504. drift/catalog gates and go test ./internal/cli pass.
Serve the versioned embedded Command Catalog (21 products / 504 tools) from
NewSchemaCommand instead of only the live tree, matching the prior branch's
release behavior and the GWS flat-leaf / Lark stable-canonical contract. Add
--all and route output through internal/output for --format/--jq/--fields.
Port schema_catalog_test.go asserting 504/21 embedded catalog integrity.
Helper subtree and live Cobra tree remain as fallbacks.
Consolidate the prior schema branch (old discovery-based architecture) into the
upstream-based dynamic-schema implementation. Merged tree keeps the upstream
static-endpoint architecture with dynamic schema; old discovery/generator/compat
packages are not carried over (incompatible with upstream, superseded by the
live-tree dynamic schema). Old schema data assets (agent metadata, destructive
safety annotations, conference metadata) remain present via the ported runtime.
Brings origin/feat/schema-gws-flat history in, so pushing is a fast-forward.
Restore dynamic dws schema on top of upstream static-endpoint runtime
(v1.0.52) without re-introducing service discovery:
- port schema runtime (runtime_schema/schema_catalog/schema_agent_metadata/
schema_hints) + embedded agent & interface metadata + ir data structures
- ir/catalog.go: drop discovery-dependent BuildCatalog, keep runtime types
- canonical.go NewSchemaCommand: build schema from the live Cobra tree via
runtimeSchemaPayload instead of the stub
- add schema_support.go and design doc docs/schema-dynamic-endpoint-design.md
go build ./... passes; go test ./... 44 packages pass (only unrelated
post-goreleaser packaging tests fail with a known tar format issue).
Add skills/mono/schema-hints/conference.json annotating all 33 conference
meeting-control tools with agent_summary, effect and reviewed=true. Mark
end-meeting-for-all as risk=high + confirmation=user_required; mute-all and
cloud-record start/stop as risk=medium.
Coverage: missing agent_summary 81->48, missing effect 173->140,
reviewed=true 4->37. Drift/catalog gates, go test and 560-case smoke pass.
* fix: honor global --jq/--fields on product commands; round-2 QA fixes
Make the global --jq / --fields output filters actually work for the
product (MCP) commands. The helper Formatter used by every product
command ignored them, so they were silent no-ops there (they already
worked for `dws api`). Expose Fields()/JQ() on the ToolCaller interface
and apply the existing output.WriteFiltered path in the helper
Formatter's PrintJSON. The handful of bespoke utility commands
(auth/config/profile/...) still encode directly and are documented as
such.
Additional CLI fixes surfaced by the second real-machine QA pass:
- sheet write-image: emit clean JSON under --format json (suppress the
progress lines that leaked onto stdout, same as media-upload/export)
- sheet range batch-set-style: under --format json, collect per-item
results into a single JSON object instead of printing N separate ones
- chat download-media: create the output directory when missing and
strip URL-encoded path separators from the inferred filename so the
file actually lands instead of failing on a missing subdirectory
- pat chmod, aitable, sheet, chat, attendance: correct --help text
(real scope names, non-existent subcommands, flag requiredness,
alxs -> axls typo)
Helper scripts (mono and multi):
- minutes_extract_todos: parse dingtalkTodoList/actions (there is no
todos key), so todos are no longer silently dropped
- sync the multi copies of chat_export_messages / chat_history_with_user
(were crashing with AttributeError), minutes_list_parse /
minutes_recent_summary, and calendar_free_slot_finder to the fixed
mono versions
Skill docs (mono and multi): correct return-structure keys, flag names,
deprecated command routing (doc download -> drive download), enum values
and server-side limitations across products; update the global
reference to note --jq/--fields now apply to product commands.
* fix(skill): make skill setup --dry-run a no-op preview; doc/help fixups
skill setup ignored the global --dry-run flag and always wrote the skill
files (overwriting an existing install). Short-circuit into a preview
that lists the source, target dirs and selected sub-skills without
touching the filesystem.
Also correct a few doc/help mismatches found in the round-3 health check:
- attendance vacation balance/records quick-reference examples were
missing the required --leave-code flag
- mail mailbox list --help described the returned field as "mailboxes"
but the real field is "emailAccounts"
* docs: clarify --fields projects top-level/list keys, use --jq for nested
* docs: drop QA voice ("真机") and don't state env-specific quirks as absolute rules
The QA-driven doc/comment edits leaked test-process narration ("真机实测")
and this environment/account's quirks stated as universal rules into the
skill files, which are general-purpose instructions for any org/account.
Strip the "真机" narration everywhere; reword environment-specific findings
(PUBLIC sharing disabled by org policy, transient 1002, sender-open-dingtalk-id
behaviour) from absolute bans into conditional hints; keep genuinely
universal command behaviour, just without the QA voice.
Fix CLI command bugs surfaced by full real-machine QA:
- aitable: make chart/dashboard share update --enabled a string flag so
"--enabled false" disables instead of silently enabling (bool flag +
space-syntax help example inverted the action); clarify chart update
requires --config; make form get filter by view-id client-side so it
returns a single form; drop inline // comments from chart JSON examples
- chat: resolve conversation-info --user to openDingTalkId, register
--id/--conversation-id/--chat aliases; cap list-all-conversations
--limit at 100 and reject larger values instead of silent truncation;
detect webhook errcode failures instead of wrapping them as success;
remove duplicate group/members subcommand registration in help
- contact: register --dept/--depts as the primary dept flags to match
the RunE parsing (were only registered as --id/--ids)
- sheet: emit clean JSON for media-upload and export under --format json,
suppressing progress lines that leaked onto stdout
- wiki: correct node create --type enum (drop unsupported asheet, add
axls/able/appt/adraw/amind)
- ding: default message list --type to ALL since the server rejects an
empty type
Fix helper scripts (mono and multi):
- aitable import/export flag names and tableId length regex
- mail search --limit, contact dept response keys and userInfo nesting
- attendance_my_record whoami compatibility, calendar_schedule_meeting
event id unwrapping, drive_tree_list recursion via fileId, report
scripts migrated off deprecated report list/detail
Sync skill docs (mono and multi) to real-machine behavior across all
products: command indexes, flag names, enums, return-structure keys, and
cross-product intent routing; annotate genuinely server-side limitations
and the no-op global --jq/--fields flags.
玉澜