plays
Usage: deepline plays [options] [command]
Search, validate, run, and manage cloud plays.
Options:
-h, --help display help for command
Commands:
resume-submission [options] <path> Inspect or explicitly resubmit the same
saved admission intent, preserving its run
ID and original revision.
check [options] <target> Check a local play file or named/prebuilt
play contract.
settings Read and change Play settings by name and
value.
trigger Read and change one Play’s cron and
webhook triggers. `plays triggers` lists
every trigger.
get [options] <target> Fetch full play details or materialize a
saved source bundle.
triggers [options] List a bounded page of triggers, or read
one trigger’s recent history.
cost [options] <play> Estimate one saved or prebuilt Play in the
current workspace.
list [options] List saved and prebuilt plays.
activity [options] Count Play runs and delivered Monitor
findings by day.
pin [options] <target> Pin an org-owned play to the top of play
listings.
unpin [options] <target> Remove an org-owned play from the pinned
group.
search [options] [task] Deprecated. Use deepline search --type
<types>.
grep [options] <query> Search active org-owned Plays through
Convex full-text search.
versions [options] List revisions for a named play.
save [options] <file> Save or update a local Play working draft
without publishing it.
deploy [options] [directory] Deploy every play in a directory as one
unit (all go live together, or none).
deployments [options] [id] Deployment history for this workspace;
pass an id for one deployment and its
plays.
publish [options] <target> Promote a saved play revision or publish a
local play file.
set-live [options] <target> Promote a saved revision or publish a
local play file.
delete [options] <target> Move an org-owned play to Trash and stop
its triggers.
restore [options] <target> Restore an org-owned play from Trash.
share Manage a play's public, unlisted share
page (advanced — exposes play code).
help [command] display help for command
Explanation:
Purpose:
Durable workflows. Prebuilt plays are Deepline-maintained waterfalls such as
name-and-domain-to-email, person-linkedin-to-email, person-to-phone, and
company search, each with a single-row and a -batch (CSV) variant. A play
you write is one TypeScript file.
When:
Any job with rows, or any job that chains providers. A play reuses finished
rows on rerun, so a failed run can be resumed instead of repaid.
The sequence:
search "<input you have -> result you need>" find the play
describe <ref> required input, CSV columns, output fields
plays cost <ref> per-row or per-run credit estimate
run <ref> --input '{"csv":"file.csv"}' --csv-out out.csv run it, save rows
Writing a play:
Start from code that already works, not a blank file.
plays get prebuilt/<name> --source --out my.play.ts
copy the closest prebuilt and edit the copy
plays check my.play.ts validate after every edit; free
run my.play.ts --input '{...}' run the local file
plays publish my.play.ts make it live for run <name> and triggers
run -e "<code>" run code inline as a play function's body
(input holds --input), like python -c
A play is a default-exported async function of its input; the file name
is its default name:
import deepline from 'deepline';
export default async function (input: { domain: string }) {
const company = await deepline.tools.execute({ id, tool, input, description });
return { company };
}
The parameter's type is the input contract; the returned object is the
result.
An optional export const config = { name, description, cron, webhook,
runtime }; at the top of the file sets the name, triggers and runtime.
--help --verbose lists the authoring rules.
Auditing plays:
plays list every saved play
plays versions --name <play> its revisions, and which one is live
plays get <play> --source --out <dir> a revision's source (--live,
--working, or --revision <n>)
plays check <file> deprecated or invalid usage, one hint per
issue; read-only and free
plays triggers every cron and webhook trigger
Cost:
search, describe, list, check, and get are free; run spends credits.
Authoring rules:
Reach the outside world only through deepline: deepline.csv,
deepline.dataset, deepline.tools.execute, deepline.runPlay, deepline.step,
deepline.fetch, deepline.secrets, deepline.customerDb, and deepline.log.
No fs, raw fetch, or shell commands, also not inside deepline.step. Local
nondeterministic work such as Date.now or Math.random goes inside
deepline.step. Call deepline.* only in the play file; pass it to helpers.
A run can be interrupted and replayed from the top; every finished step
and tool call then returns its saved result. So the play name, each
dataset key, step id, fetch key, and tools.execute id is a fixed string
literal, and renaming one pays for that work again.
Rows: deepline.dataset('key', rows).withColumn('name', (row) => ...)
.run({ key: 'domain' }); call deepline.* inside the column too. Use a business identifier as the row key. Put
one paid call in a column, and shaping, status, and miss_reason in
separate plain columns after it. Return the dataset for anything that
should export as CSV.
Dataset handles are async: await rows.count(), rows.first(),
rows.peek(n), rows.materialize(n), or for await. Array methods on a
handle fail plays check.
Tool calls: deepline.tools.execute({ id, tool, input, description,
staleAfterSeconds }). In authoring edition 11, choose freshness explicitly:
0 for fresh logical work, a positive integer for the maximum reusable
result age in seconds, or null for indefinite result reuse. A matching
completed operation receipt can still recover that operation after resume.
Read
results the way describe lists them:
result.fields.<name>.value for a value whose meaning is known, also by
field id (result.fields.person.email.value). .value throws when there are
several; .values is always an array;
result.extractedValues.<name>.get() for other declared values. Anything
else is in result.response, the provider's response with its nesting kept.
For a list, result.rows() gives one record per item, in the same format
for every tool: row.fields.email.value, and row.response for the rest.
Name the list when the tool declares several: result.rows('people').
Reuse before rebuilding: deepline.runPlay(key, 'prebuilt/<name>', input,
{ description }) runs a
prebuilt as one stage. Independent calls run together with Promise.all;
a call that needs another's result waits for it.
Failures: catch only ProviderTransientError, and only when another
provider can answer the same question. Keep the last provider outside
the catch so an exhausted waterfall fails loudly. Never branch on an
error message.
External APIs: deepline.fetch with a fixed key, and credentials through
deepline.secrets.bearer(...) or deepline.secrets.header(...), stored
with deepline secrets set. res.json is a property, not a method; check
res.ok. A play cannot loop deepline.fetch over pages: fan out over rows
with deepline.dataset, or
call one endpoint that returns the whole result.
Triggers: export const config = { cron: { schedule, timezone, input } };
or { webhook: {} }. A cron run receives {} unless cron.input is written
inline; it is pinned with the published revision.
Runtime: The default Play runtime is 30 minutes. Standard uses 1 GiB / 1 CPU.
Set runtime.size to 'standard' or '1gb' (1 GiB / 1 CPU), 'large' or '4gb'
(4 GiB / 2 CPU), or 'xlarge' or '8gb' (8 GiB / 4 CPU).
Each alias selects the same profile and price as its named size.
Compute pricing follows the workspace plan.
Use deepline billing quote for the session ceiling.
Set runtime.timeout to extend a run up to 4h.
For large datasets, keep working sets bounded and checkpoint progress;
inline ctx.runPlay shares parent memory, while ctx.runPlayAsync starts an
independent run. A larger profile does not make a job resumable. This is
not the CLI wait timeout.
Recurring sourcing: .run({ key: 'domain', mode: 'net_new' }) returns only
rows the table has not seen before.
Results: project each answer into flat columns (the value, its source,
whether it was validated, the outcome, the reason) and keep evidence
under _dl_meta, including result._metadata.execution.receiptKey. Unknown
stays null, not false.
Diagram: a /** @mermaid */ block at the top of the play draws it in the
dashboard, with // @mermaid-node <id> out:"<identifier>" above the
statement each box stands for. plays check validates it.
plays resume-submission
Usage: deepline plays resume-submission [options] <path>
Inspect or explicitly resubmit the same saved admission intent, preserving its
run ID and original revision.
Options:
--inspect Read the local submission identity only; no API calls or execution
--json Emit the identity or admission response as JSON
-h, --help display help for command
plays check
Usage: deepline plays check [options] <target>
Check a local play file or named/prebuilt play contract.
Options:
--full Include compiler diagnostics
--json [fields] Emit JSON, optionally selecting comma-separated dotted fields
-h, --help display help for command
Explanation:
Purpose:
Validate a play contract without starting a run: a local .play.ts file is
bundled and checked, a named or prebuilt play is confirmed discoverable.
When:
Before running a local play file, and after editing one. Free, so run it
every time.
Input:
A local path ending in .play.ts, or a play reference. --json prints JSON.
Output (JSON):
valid (boolean), target, name, reference, inputSchema, outputSchema,
runtimeLimit, note, and for local files any errors or warnings with the
repair steps.
Cost:
Free.
Avoid:
valid true does not mean the input you plan to pass is valid; describe
shows the required fields.
plays settings
Usage: deepline plays settings [options] [command]
Read and change Play settings by name and value.
Options:
-h, --help display help for command
Commands:
describe [options] List setting names, values, examples and defaults; no
authentication required.
get [options] Read all settings, or one setting selected with --name.
set [options] Set one setting without creating or publishing a
revision.
reset [options] Reset one setting to its schema default.
paths [options] List existing locations (one bounded page).
move-folder [options] Move or rename a folder. Every Play below it moves and
keeps its subfolders. Retry-safe: rerun to finish an
interrupted move.
help [command] display help for command
plays settings describe
Usage: deepline plays settings describe [--name <setting>] [options]
List setting names, values, examples and defaults; no authentication required.
Options:
--name <setting> Describe one setting. Available: path, displayName
--json Emit JSON
-h, --help display help for command
plays settings get
Usage: deepline plays settings get --play <play> [--name <setting>] [options]
Read all settings, or one setting selected with --name.
Options:
--play <play> Play to read
--name <setting> Read one setting. Available: path, displayName
--json Emit JSON
-h, --help display help for command
plays settings set
Usage: deepline plays settings set --play <play> --name <setting> --value <value> [options]
Set one setting without creating or publishing a revision.
Options:
--play <play> Play to update
--name <setting> Setting name. Available: path, displayName
--value <value> New value; see settings describe --name <setting>
--dry-run Show the patch without changing settings
--json Emit JSON
-h, --help display help for command
plays settings reset
Usage: deepline plays settings reset --play <play> --name <setting> [options]
Reset one setting to its schema default.
Options:
--play <play> Play to update
--name <setting> Setting name. Available: path, displayName
--dry-run Show the patch without changing settings
--json Emit JSON
-h, --help display help for command
plays settings paths
Usage: deepline plays settings paths [options]
List existing locations (one bounded page).
Options:
--root <person|team> Filter by root
--limit <n> Plays to inspect, 1–100 (default: "50")
--cursor <cursor> Continue from a previous page
--json Emit JSON
-h, --help display help for command
plays settings move-folder
Usage: deepline plays settings move-folder --from <path> --to <path> [--owner <user-id>] [options]
Move or rename a folder. Every Play below it moves and keeps its subfolders.
Retry-safe: rerun to finish an interrupted move.
Options:
--from <path> Folder to move, e.g. team/Research
--to <path> Its new path including its name, e.g. team/Archive/Research
--owner <user-id> Personal-root owner, when several people have the --from
folder or --to is person
--dry-run Show the planned moves without changing settings
--json Emit JSON
-h, --help display help for command
plays trigger
Usage: deepline plays trigger [options] [command]
Read and change one Play’s cron and webhook triggers. `plays triggers` lists
every trigger.
Options:
-h, --help display help for command
Commands:
get [options] Show the cron and webhook triggers of a
Play.
set-cron|set-schedule [options] Schedule the Play (replaces any existing
schedule).
clear-cron|clear-schedule [options] Remove the schedule.
set-webhook [options] Turn on the webhook trigger. Without
--auth, keeps the current authentication.
clear-webhook [options] Remove the webhook trigger and retire its
URL.
help [command] display help for command
plays trigger get
Usage: deepline plays trigger get [options]
Show the cron and webhook triggers of a Play.
Options:
--play <play> Play to read
--json Emit JSON
-h, --help display help for command
plays trigger set-cron
Usage: deepline plays trigger set-cron|set-schedule [options]
Schedule the Play (replaces any existing schedule).
Options:
--schedule <cron> 5-field cron, e.g. "0 9 * * 1-5"
--timezone <iana> IANA timezone; default UTC
--input <json> Static JSON object passed to every cron run
--play <play> Play to change
--expected-revision <id> Refuse the change if the live revision is no longer
this one
--dry-run Show the change without publishing
--json Emit JSON
-h, --help display help for command
plays trigger clear-cron
Usage: deepline plays trigger clear-cron|clear-schedule [options]
Remove the schedule.
Options:
--play <play> Play to change
--expected-revision <id> Refuse the change if the live revision is no longer
this one
--dry-run Show the change without publishing
--json Emit JSON
-h, --help display help for command
plays trigger set-webhook
Usage: deepline plays trigger set-webhook [options]
Turn on the webhook trigger. Without --auth, keeps the current authentication.
Options:
--auth <mode> none, standard (Standard Webhooks), svix, or hmac
(legacy sha256)
--secret <name> Deepline Secret name; repeat for rotation
(standard/svix)
--header <name> Signature header (hmac)
--signature-prefix <prefix> Signature prefix (hmac)
--tolerance-seconds <n> Replay window in seconds (standard/svix), default
300
--play <play> Play to change
--expected-revision <id> Refuse the change if the live revision is no
longer this one
--dry-run Show the change without publishing
--json Emit JSON
-h, --help display help for command
plays trigger clear-webhook
Usage: deepline plays trigger clear-webhook [options]
Remove the webhook trigger and retire its URL.
Options:
--play <play> Play to change
--expected-revision <id> Refuse the change if the live revision is no longer
this one
--dry-run Show the change without publishing
--json Emit JSON
-h, --help display help for command
plays get
Usage: deepline plays get [options] <target>
Fetch full play details or materialize a saved source bundle.
Options:
--json Emit JSON output
--source Read source; prints single-file Plays or combines with
--out to write the full bundle
--out <path> Write source to a file or, for multi-file Plays, a
directory
--working Read the current working draft source (default)
--live Read the live revision source
--revision <number> Read a specific revision source
-h, --help display help for command
Explanation:
Purpose:
Full metadata for a play: revisions, source, triggers, and latest runs.
When:
Only when describe is not enough: reading the source (--source --out
file.play.ts), checking revisions, or auditing a saved play.
Input:
The play reference. --source prints or writes the source; --working,
--live, --revision <n> pick which source.
Output (JSON):
The play record with revision and trigger metadata; with --source, the
TypeScript source.
Cost:
Free.
plays triggers
Usage: deepline plays triggers [options]
List a bounded page of triggers, or read one trigger’s recent history.
Options:
--cursor <cursor> Continue the trigger inventory
--binding <id> Read history for this trigger binding
--history-limit <count> Recent history cap, 1–30 (default: "10")
--json Emit JSON output
-h, --help display help for command
plays cost
Usage: deepline plays cost [options] <play>
Estimate one saved or prebuilt Play in the current workspace.
Options:
--json Emit the cost estimate as JSON
-h, --help display help for command
Explanation:
Purpose:
Read the advisory cost estimate for one saved or prebuilt Play.
Input:
<play> is a bare name or qualified reference. --json prints the exact result.
Output:
reference, costEstimate, and for prebuilts observedCost. Estimates include
the output unit, observed or static range, sample size and compute ceiling.
Cost:
Read-only. Uses the current workspace's pricing; does not start a run.
plays list
Usage: deepline plays list [options]
List saved and prebuilt plays.
Options:
--limit <count> Page size
--cursor <cursor> Continue a previous page
--show-cost Include the per-row or per-run cost estimate
--summary Load metadata without authored source or revision
payloads
--json Emit JSON output
--categories <categories> Filter by comma-separated canonical tool categories
--tag <tag...> Filter by the job the play does: smb,
find-prospects, qualify, contact-info, signals
-h, --help display help for command
Explanation:
Purpose:
Inventory of saved and prebuilt plays, one bounded page at a time.
Add --show-cost for estimates across the page.
When:
To price a catalog page: plays list --show-cost, then read the
cost estimate line for the play, or in JSON observedCost (prebuilts) or
costEstimate.
Input:
--show-cost adds estimates. --summary loads metadata without source or
revision payloads. --limit and --cursor page. --categories
filters by tool category. --json prints JSON.
Output (JSON):
plays[] with playKey, reference, displayName, description, origin,
runCount, lastRun, and with --show-cost: costEstimate {status, unit
("row" or "run"), lowCredits, highCredits, source, sampleRunCount} or
{status: "unavailable", reason, message}, and for prebuilts also
observedCost, the same object search returns; count, has_more, next_cursor,
next (the next-page command), render.sections with a "cost estimate:"
line per play.
Cost:
Free.
Avoid:
For a prebuilt, search and describe show the same observedCost; for your
own plays only this command estimates cost.
Pricing a run:
Multiply lowCredits..highCredits (observedCost: avgCredits, or the typical
range before runs are observed) by the row count when unit is "row", or
use it directly when unit is "run". Compare with the billing balance.
When the estimate is unavailable, run a two-row slice first and
read its actual spend with billing usage --run-id <run-id>.
Output fields:
costEstimate.source "observed" (from past runs; shown with ~) or
"static" (a hand-written range or the play's
declared tools).
costEstimate.sampleRunCount how many runs the observation is based on.
observedCost avgCredits, medianCredits, maxCredits per unit across
all workspaces' runs, or status "unavailable" with
typical {lowCredits, highCredits, unit}.
has_more / next_cursor the page is bounded; pass --cursor to continue.
is_truncated true when the page was cut for size.
plays activity
Usage: deepline plays activity [options]
Count Play runs and delivered Monitor findings by day.
Options:
--time-zone <iana> Calendar timezone (default: UTC) (default: "UTC")
--days <count> Trailing calendar days, 1–45 (default: "30")
--end-date <yyyy-mm-dd> Last local calendar day
--group-by <kind> kind or trigger_source (default: "kind")
--json Emit exact counts and coverage as JSON
-h, --help display help for command
plays pin
Usage: deepline plays pin [options] <target>
Pin an org-owned play to the top of play listings.
Options:
--dry-run Print the planned change without saving it
--json Emit JSON output
-h, --help display help for command
plays unpin
Usage: deepline plays unpin [options] <target>
Remove an org-owned play from the pinned group.
Options:
--dry-run Print the planned change without saving it
--json Emit JSON output
-h, --help display help for command
plays search
Usage: deepline plays search [options] [task]
Deprecated. Use deepline search --type <types>.
Options:
--all Include plays your workspace created
--compact Accepted for compatibility; results are compact
--categories <categories> Only tools in these categories, comma-separated
--limit <count> Results per page, 1-50
--offset <count> Results to skip, for the next page
--summary-width <chars> Description characters per result (default 140)
--json Print JSON instead of the readable list
-h, --help display help for command
plays grep
Usage: deepline plays grep [options] <query>
Search active org-owned Plays through Convex full-text search.
Options:
--json Emit JSON output
-h, --help display help for command
Explanation:
Purpose:
Full-text search over plays this workspace authored (not prebuilt ones).
When:
Finding a saved play by a phrase in its source or description.
Input:
A query. --json prints JSON.
Output (JSON):
Matching owned plays with the matched fields and terms.
Cost:
Free.
plays versions
Usage: deepline plays versions [options]
List revisions for a named play.
Options:
--name <name> Saved play name
--json Emit JSON output
-h, --help display help for command
Explanation:
Purpose:
List the saved revisions of a workspace play, newest first, and which one
is live.
When:
Before making an older revision live or running a specific revision.
Input:
--name <play>. --json forces JSON.
Output (JSON):
The play's revisions, each with its revision id and whether it is live.
Cost:
Free.
plays save
Usage: deepline plays save [options] <file>
Save or update a local Play working draft without publishing it.
Options:
--expected-artifact <hash> Require the checked artifact hash
--dry-run Bundle and show the planned draft save without
mutation
--json Emit JSON output
-h, --help display help for command
plays deploy
Usage: deepline plays deploy [options] [directory]
Deploy every play in a directory as one unit (all go live together, or none).
Options:
--manifest <file> Manifest file name in the directory (default:
deepline.json)
--repo <owner/name> Repository to record on the deployment
--ref <ref> Branch or ref to record on the deployment
--commit <sha> Commit to record on the deployment
--json Emit JSON output
-h, --help display help for command
plays deployments
Usage: deepline plays deployments [options] [id]
Deployment history for this workspace; pass an id for one deployment and its
plays.
Options:
--limit <n> Max rows (default 50)
--json Emit JSON output
-h, --help display help for command
plays publish
Usage: deepline plays publish [options] <target>
Promote a saved play revision or publish a local play file.
Options:
--latest Promote the newest saved revision
--revision-id <id> Revision to promote
--expected-artifact <hash> Require the local file to produce this checked
artifact hash
--dry-run Bundle and show the planned publish without
mutation
--json Emit JSON output
-h, --help display help for command
Explanation:
Purpose:
Publish a local .play.ts file or promote a saved revision, so run
<name>, triggers, and deepline.runPlay callers use it. The same operation as
plays set-live.
When:
Once a play run by file path gives the right result. Running a local file
never publishes it.
Input:
A local path ending in .play.ts, or a saved play name with --latest or
--revision-id <id>. --dry-run bundles and prints the planned artifact
without publishing. --expected-artifact <hash> refuses to publish when the
file no longer produces the artifact hash plays check printed.
Output (JSON):
success, name, liveVersion, revisionId, artifactHash, and the trigger
metadata and bindings now live.
Cost:
Free. It changes cloud state: every caller of the play name gets the new
version at once.
Avoid:
A cron or webhook trigger cannot report its own failure. After publishing
a triggered play, check that a failure notifies someone: notifications
--help.
plays set-live
Usage: deepline plays set-live [options] <target>
Promote a saved revision or publish a local play file.
Options:
--latest Promote the newest saved revision
--revision-id <id> Revision to promote
--expected-artifact <hash> Require the local file to produce this checked
artifact hash
--dry-run Bundle and show the planned publish without
mutation
--json Emit JSON output
-h, --help display help for command
Explanation:
Purpose:
Make one version of a play live, so run <name>, triggers, and
deepline.runPlay callers use it. A local .play.ts file is bundled, validated,
saved as a revision, and promoted in one step; a saved play promotes one
of its existing revisions.
When:
After iterating on a local file with run ./file.play.ts, once its
result is right. Running a local file never publishes it.
Input:
A local path ending in .play.ts, or a saved play name with --latest or
--revision-id <id>. --dry-run bundles and prints the planned artifact
without publishing. --expected-artifact <hash> refuses to publish when the
file no longer produces the artifact hash plays check printed.
Output (JSON):
success, name, liveVersion, revisionId, artifactHash, previousLiveVersion,
previousArtifactHash, and the trigger metadata and bindings now live.
Cost:
Free. It changes cloud state: every caller of the play name gets the new
version at once.
Avoid:
Do not publish to test a change; run the file by path instead.
A cron or webhook trigger cannot report its own failure. After publishing
a triggered play, check that a failure notifies someone: notifications
--help.
plays delete
Usage: deepline plays delete [options] <target>
Move an org-owned play to Trash and stop its triggers.
Options:
-y, --yes Confirm deletion
--dry-run Show the planned archive without mutating the play
--json Emit JSON output
-h, --help display help for command
Explanation:
Purpose:
Move a workspace play to Trash and stop its triggers. Its revisions and
run history are kept, so it can be restored.
When:
When the user asks to remove a play. Prebuilt plays cannot be deleted.
Input:
The play name. --yes confirms; --dry-run shows the planned change.
Output (JSON):
The deleted play and the triggers that were stopped.
Cost:
Free.
plays restore
Usage: deepline plays restore [options] <target>
Restore an org-owned play from Trash.
Options:
--json Emit JSON output
-h, --help display help for command
Explanation:
Purpose:
Restore a play from Trash. The saved play is enabled again and its trigger
bindings are recreated.
When:
After a play was deleted by mistake.
Input:
The play name. --json forces JSON.
Output (JSON):
The restored play.
Cost:
Free.
plays share
Usage: deepline plays share [options] [command]
Manage a play's public, unlisted share page (advanced — exposes play code).
Options:
-h, --help display help for command
Commands:
status [options] <play> Show a play's share-page status and revision
picker.
publish [options] <play> Publish/repoint a PUBLIC page to a revision.
Advanced; not recommended unless explicitly
requested and risks confirmed.
update [options] <play> Update share-page settings without moving the
pointer.
regenerate [options] <play> Regenerate the public page's LLM copy for a
revision.
unpublish [options] <play> Unshare: permanently delete the public page
(frees it). Requires --yes.
help [command] display help for command
plays share status
Usage: deepline plays share status [options] <play>
Show a play's share-page status and revision picker.
Options:
--json Emit JSON output
-h, --help display help for command
plays share publish
Usage: deepline plays share publish [options] <play>
Publish/repoint a PUBLIC page to a revision. Advanced; not recommended unless
explicitly requested and risks confirmed.
Options:
--latest Publish the newest revision (default).
--version <n> Publish a specific version number.
--revision-id <id> Publish a specific revision id.
--seo <mode> SEO indexing: index | noindex.
--show-cost Show average Deepline credit cost on the page.
--show-latency Show average latency on the page.
--no-run-check Skip the "must have a completed run" gate.
-y, --yes Confirm public exposure of the play code/structure
(required).
--dry-run Show the planned publish without writing.
--json Emit JSON output
-h, --help display help for command
plays share update
Usage: deepline plays share update [options] <play>
Update share-page settings without moving the pointer.
Options:
--seo <mode> SEO indexing: index | noindex.
--show-cost Show average Deepline credit cost.
--hide-cost Hide average Deepline credit cost.
--show-latency Show average latency.
--hide-latency Hide average latency.
--json Emit JSON output
-h, --help display help for command
plays share regenerate
Usage: deepline plays share regenerate [options] <play>
Regenerate the public page's LLM copy for a revision.
Options:
--version <n> Regenerate a specific version number.
--revision-id <id> Regenerate a specific revision id.
--json Emit JSON output
-h, --help display help for command
plays share unpublish
Usage: deepline plays share unpublish [options] <play>
Unshare: permanently delete the public page (frees it). Requires --yes.
Options:
-y, --yes Confirm taking the public page down (required).
--json Emit JSON output
-h, --help display help for command