@cloudglue/tinycloud 0.3.10 → 0.3.12

package/README.md

@@ -17,11 +17,11 @@ The npm package is a small launcher: on first run it downloads the matching
 platform distribution from Cloudglue's CDN (cached under
 `~/.tinycloud/versions/<version>/`), verifies its checksum, and execs the real
 binary. The package version pins the binary version, so
-`npx @cloudglue/tinycloud@0.3.10` always runs tinycloud 0.3.10. It also adds two
+`npx @cloudglue/tinycloud@0.3.12` always runs tinycloud 0.3.12. It also adds two
 wrapper commands:
 
 ```bash
-tinycloud install --version 0.3.10   # pre-download a version
+tinycloud install --version 0.3.12   # pre-download a version
 tinycloud install --latest          # install latest stable and pin to it
 tinycloud update                    # move to latest stable, prune old versions
 ```

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@cloudglue/tinycloud",
-  "version": "0.3.10",
+  "version": "0.3.12",
   "description": "Agent CLI for deep video and image work, by Cloudglue. Downloads the tinycloud binary on first run.",
   "bin": {
     "tinycloud": "bin/tinycloud.js"

package/skills/tinycloud/reference/glossary.md

@@ -39,10 +39,11 @@ connector?" or an envelope field needs explaining.
   `library collections show <col>` until every `files[].status` is `completed`
   before querying.
 - **Data connector** — a linked external source of recordings (Zoom, Grain,
-  Google Drive, Dropbox, Loom, S3/GCS). `tinycloud library connectors …`
-  lists, browses (`files`, with provider-specific filters), and syncs
-  individual items by URI (e.g. `grain://recording/<id>`) so they become
-  Cloudglue files.
+  Gong, Recall, Google Drive, Dropbox, S3/GCS). `tinycloud library connectors …`
+  lists, browses (`files`, with provider-specific filters; rows carry provider
+  metadata on 0.3.11+), peeks one item's provider metadata without syncing
+  (`inspect`, 0.3.11+), and syncs individual items by URI
+  (e.g. `grain://recording/<id>`) so they become Cloudglue files.
 - **Source** — anything a verb accepts as input: a local path, URL,
   `cloudglue://files/<id>` URI, connector URI, collection, or a bare file-id
   UUID (normalized to `cloudglue://files/<id>`; an existing local path of the
@@ -70,7 +71,8 @@ connector?" or an envelope field needs explaining.
   (`status`, `data`, `ref`, `meta`, `error`, …). The machine contract; see
   reference/envelope.md.
 - **Watch context / describe** — the reusable analysis `tinycloud watch`
-  produces (summary, segments, transcript-ish context). Cached locally and
+  produces (summary, segments with their spoken utterances inlined as
+  `speech[]`, transcript-ish context). Cached locally and
   mirrored in Cloudglue, so later `extract`/`ask`/`search` reuse it instead
   of re-analyzing.
 - **See / image describe (0.3.7+)** — the image counterpart of `watch`:

package/skills/tinycloud/reference/verbs.md

@@ -54,7 +54,7 @@ Flags shared by most verbs are listed once at the bottom.
 tinycloud watch <source> [--segment uniform:20|chapters|shots|segments]
   [--shot-min-seconds <s>] [--shot-max-seconds <s>]
   [--profile default|light|custom] [--speech-only | --visual-only]
-  [--start <t>] [--end <t>] [--transcript] [--content] [--json-index]
+  [--start <t>] [--end <t>] [--transcript] [--content]
   [--background]
 ```
 
@@ -65,6 +65,20 @@ bounds are part of the cache key, so tuned and default shot passes never
 collide. `watch` is **video/audio only** — point it at an image and it errors
 ("watch analyzes video/audio; for an image use `tinycloud see`").
 
+Segments carry the spoken words (0.3.12+): each segment (and embedded shot)
+inlines its overlapping utterances as `speech: string[]`. A source with
+speech but no visual segmentation — an audio file, `--speech-only`, a very
+short or windowed source — gets uniform 20s speech segments synthesized
+(`"segmentation": "uniform:20"`) instead of an empty `segments` list. Caches
+written by older versions heal automatically on the next watch or search.
+
+`--transcript` adds the full transcript as `data.transcript` (speaker-labeled
+when more than one speaker); `--content` inlines the complete describe
+markdown as `data.content`. Both are derived per invocation — they don't
+change the cache key and are served on cache hits, so asking for them never
+re-runs an analysis. `--json-index` was removed in 0.3.12 (it had long been a
+no-op; passing it now errors as an unknown flag).
+
 ### see — analyze an image (cloud, 0.3.7+)
 
 ```bash
@@ -122,6 +136,9 @@ tinycloud search "<keyword>" [--in <paths|source-ids|collection-ids|all>]
 
 Searches cached context locally, including cached `see` image results (0.3.7+),
 so a describe you've already run is greppable without another cloud call.
+`--field speech` matches the utterances inlined on cached watch segments —
+including speech-only/audio watches (0.3.12+; older cache entries heal in
+place during the scan, nothing is re-billed).
 
 ### probe — semantic search (cloud)
 
@@ -206,6 +223,7 @@ tinycloud library collections delete <col_id> --json
 tinycloud library collections entities <col_id> <source> [--limit <n>] [--offset <n>] --json   # read a video's entities
 tinycloud library connectors list --json
 tinycloud library connectors files <connector-id> [--limit 25] [--page-token <t>] --json
+tinycloud library connectors inspect [<connector-id>] <uri-or-share-link> --json   # metadata peek, no file created (0.3.11+)
 tinycloud library connectors sync [<connector-id>] <uri-share-link-or-public-url> --json
 ```
 
@@ -250,17 +268,32 @@ Dropbox file share links sync server-side via the connector's OAuth
 (including login-gated links); `zoom.us/rec/share` links resolve best-effort
 (Zoom mints a new token per copy — the recording-detail link is the reliable
 form). Link warnings are advisory and surface in `data.warnings` rather than
-blocking the sync. Non-connector public URLs (direct media URLs, TikTok,
+blocking the sync. Synced files carry provider `source_metadata` (title,
+participants, summary) for Grain, Zoom, Recall, Google Drive, Dropbox, and
+Gong. Non-connector public URLs (direct media URLs, TikTok,
 Loom, public Dropbox links without a connector) sync into a standalone
 Cloudglue file via direct URL ingestion — same command, no connector needed.
 YouTube URLs cannot sync; use `tinycloud grab` instead.
 
 `connectors files` also takes provider-specific filters: `--from`/`--to`
-(Zoom, Grain dates), `--folder-id` (Google Drive), `--path` (Dropbox),
-`--bucket`/`--prefix` (S3/GCS), `--title-search`/`--team`/`--meeting-type`
-(Grain). Collection IDs (`col_…`) are stable; collection names are
+(dates — every provider except S3/GCS; Zoom and Gong default to the last
+6 months), `--folder-id` (Google Drive), `--path` (Dropbox),
+`--bucket`/`--prefix` (S3/GCS — bucket required), `--title-search` (Grain,
+Zoom, Google Drive, Dropbox, Gong), `--team`/`--meeting-type` (Grain).
+Filters a provider can't honor are ignored, and filtered pages can come back
+short or even empty while more remain — keep paging until `next_page_token`
+is null. On 0.3.11+ each row also carries provider `metadata` (participants,
+host, duration, AI summary) when the source exposes it — use it to pick files
+before syncing. Collection IDs (`col_…`) are stable; collection names are
 display-only.
 
+`connectors inspect <uri>` (0.3.11+, feature `library.connectors.inspect.v1`)
+returns one item's provider `source_metadata` WITHOUT materializing a file —
+same URI/share-link forms as sync, connector id optional. S3/GCS objects have
+nothing richer to inspect. On older binaries it fails with an unknown-command
+error — fall back to `connectors sync` (idempotent) and read
+`source_metadata` from the sync envelope instead.
+
 ### jobs — async work
 
 ```bash

package/skills/tinycloud/scripts/preflight.sh

@@ -9,11 +9,11 @@ set -u
 
 # Mirror tinycloud-skill.json: min_version / supported_range upper bound
 # (CI diffs these against the manifest).
-MIN_VERSION="0.3.7"
+MIN_VERSION="0.3.12"
 MAX_VERSION_EXCLUSIVE="0.4.0"
 INSTALL_CMD='curl -fsSL https://app.cloudglue.dev/tinycloud.sh | bash'
 # Kept in sync with ../tinycloud-skill.json required_features (CI diffs them).
-REQUIRED_FEATURES="envelope.v1 watch.v1 see.v1 extract.v1 extract.images.v1 caption.v1 search.v1 probe.v1 ask.v1 clip.v1 grab.v1 face.v1 jobs.v1 library.collections.v1 library.collections.create.v1 library.collections.mutate.v1 library.collections.entities.v1 library.sync.url.v1 workflow.v1 publish.v1 publish.manage.v1 publish.video.v1 setup.v1"
+REQUIRED_FEATURES="envelope.v1 watch.v1 watch.speech.v1 see.v1 extract.v1 extract.images.v1 caption.v1 search.v1 probe.v1 ask.v1 clip.v1 grab.v1 face.v1 jobs.v1 library.collections.v1 library.collections.create.v1 library.collections.mutate.v1 library.collections.entities.v1 library.sync.url.v1 workflow.v1 publish.v1 publish.manage.v1 publish.video.v1 setup.v1"
 
 # 1) Binary present and responsive?
 if ! command -v tinycloud >/dev/null 2>&1; then

package/skills/tinycloud/tinycloud-skill.json

@@ -1,11 +1,12 @@
 {
-  "skill_version": "0.3.10",
+  "skill_version": "0.3.12",
   "tinycloud": {
-    "min_version": "0.3.7",
-    "supported_range": ">=0.3.7 <0.4.0",
+    "min_version": "0.3.12",
+    "supported_range": ">=0.3.12 <0.4.0",
     "required_features": [
       "envelope.v1",
       "watch.v1",
+      "watch.speech.v1",
       "see.v1",
       "extract.v1",
       "extract.images.v1",