agentsclimarketplace
Sign in

Apify actor task

Skill MrBridgeHQ/apify-actor-task-claude/skills/apify-actor-task

Use when planning, authoring, or auditing the publication of an Apify Actor task as a public use-case landing page - deciding which tasks to publish (one keyword-targeted use-case each), producing the publication spec (slug, SEO title, SEO description, which input fields to display, dataset view, prefilled input JSON) for the Console Publication tab, or auditing already-published tasks. Triggers on "publish an Apify task", "Actor task landing page", "Examples tab", "task SEO title / slug / description", "which tasks should I publish", "publish-task", EN or FR. Task publishing is an experimental Apify feature (contact Apify support for access). Distinct from the Actor README/Store content, pricing, and prose humanization.From its SKILL.md

Install
npx -y skills add MrBridgeHQ/apify-actor-task-claude --skill apify-actor-task

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

7.6 KB, ~1.6k tokens by cl100k_base, as published. Nobody here has run it

Apify Actor Task

This skill owns task publication for Apify Actors: turning a saved task into a public, keyword-targeted use-case landing page via the Console Publication tab → Examples tab. It covers task selection strategy, the full publication spec (slug, SEO title, SEO description, displayed input fields, dataset view, prefilled input JSON), and post-publish audits. Fine SEO-copy doctrine (length formulas, wording patterns) is delegated to a dedicated Store-content authoring skill, not re-derived here.

Experimental feature. Task publishing via the Apify Console Publication tab is not available to all users - it is a gated experimental feature. If the Console Publication tab is not visible on your Actor, contact Apify support to request access before proceeding with any step below.

First action when invoked

  1. Detect the mode (PLAN / WRITE / AUDIT) from the request. If ambiguous, ask once.
  2. Load the matching reference(s) from the table below - do not produce output from memory alone.
  3. Then produce the artifact for that mode.

Never skip step 2. The references are the source of truth for validation rules, field specs, and canonical examples.

Load-on-demand table

Mode / needLoad this
PLAN - which tasks to publish, how many, naming strategyreferences/task-strategy.md
WRITE - produce a publication spec for one taskreferences/publication-spec.md (+ delegate fine copy to your Store-content authoring skill)
AUDIT - review an existing published taskreferences/publication-spec.md (the validation checklist section)
How-to / Console mechanics - where to click, Publication tab fields, dataset viewsreferences/console-howto.md

Delegation table

This skill owns task selection and the publication spec. It does not re-derive copy doctrine or schema authoring - those belong to sister skills/tooling in your own setup.

TriggerRedirect
Fine SEO title/description wording (length formulas, patterns)Your Store-content authoring skill (SEO-description doctrine)
Keyword research for a task's target use-caseYour Store-content authoring skill (keyword + GEO/AEO doctrine)
Humanizing the title/description proseYour prose-humanization skill
Defining or restructuring the dataset view itselfYour Store-content authoring skill (dataset-schema doctrine)
Marking an input field isSecret in the input schemaYour Store-content authoring skill (input-schema doctrine)
Pricing/monetization of the underlying ActorYour Actor-monetization skill

Rules

R1 - Affiliate parameter. If you run an Apify affiliate funnel, every apify.com/… URL produced by this skill should carry your affiliate parameter (?fpr=YOUR_AFFILIATE_ID). Apply it consistently or not at all. The examples below use the placeholder YOUR_AFFILIATE_ID - substitute your own.

R2 - Brand name, no hyperlink. A task landing page has no free prose body, so there is no place to embed a hyperlink. Your brand name MAY be cited in plain text (no hyperlink) in the SEO description when the use-case naturally allows it (e.g. "by <your brand>"). Do not force it.

R3 - Experimental gate. Before any WRITE or AUDIT step, confirm the user has the Publication tab enabled. If uncertain, instruct them to contact Apify support first.

R4 - Examples are illustrative. All worked examples use a generic, fictional Actor (a Vivino wine-data scraper) purely to illustrate the technique. Replace it with your own Actor and use-cases.

R5 - No internal business context. Never include any internal business context (margins, client names, org structure, private analytics) in a published task spec. A plain-text brand name in the SEO description is fine (see R2 - no hyperlink); business internals are not.

R6 - Secret-field gate before publication. Before recommending that a task be published, confirm every credential or sensitive input field has isSecret: true in the Actor's input schema. Apify auto-masks isSecret fields in the public task view. The AUDIT mode must check this explicitly.

Anti-patterns (doctrine counters)

Failure modeCorrect doctrine
Conflating the task slug with the SEO titleThey are different fields - never treat one as the other.
Inventing platform mechanics (e.g. "Featured task slots", "publish #1-3 as Featured")No such slots exist. There is one Publication tab per task. Do not invent mechanics not documented in references/console-howto.md.
Treating the public input-field list as the task's full inputThe displayed fields are a selection for the UI; the task still runs with its full input configuration. Hiding a field from the display does not remove it from the run.
Fixing a leaked secret by removing the field from the input schemaFix it with isSecret: true on the existing field (R6). The public display list is a selection, not a schema edit.

Field-level specs (slug format, SEO title/description length & formulas, goal-framed title rules) live in references/publication-spec.md and are delegated to your Store-content authoring skill's SEO-description doctrine - not restated here.

When to ask / when to defer

Ask the user when:

  • Mode (PLAN / WRITE / AUDIT) is ambiguous.
  • The target Actor and its existing input schema are not provided (needed for WRITE and AUDIT).
  • The desired use-case keyword for a new task is unclear.

Defer to the user when:

  • Choosing which specific keyword to target (competitive positioning decision).
  • Approving the final prefilled input JSON values (user knows their Actor's realistic defaults).
  • Deciding whether to request the experimental feature from Apify support (user action required).

Available references

FilePurpose
references/task-strategy.mdHow to select, scope, and name tasks - one use-case per task, goal-framed titles, how many to publish, slug conventions
references/publication-spec.mdFull spec for the Console Publication tab - all fields, validation checklist, display-field selection, dataset-view selection, prefilled JSON, AUDIT checklist
references/console-howto.mdStep-by-step Console walkthrough - where the Publication tab lives, how to set each field, how to verify the published Examples tab

What this skill is not

  • Not the Actor README or Store description. Publishing artifacts (Store description, README, input/output schemas) belong to a dedicated Store-content authoring skill.
  • Not pricing or monetization. PPE setup, Actor.charge(), revenue audit belong to a dedicated Actor-monetization skill.
  • Not creating or saving the task input itself. Saving a task configuration in the Apify Console is a basic product operation, not a skill concern.
  • Not a non-Apify distribution channel. Listing on third-party directories belongs to a dedicated distribution skill.

What ships with it: 6 files

40.7 KB alongside SKILL.md

assets/

references/

Keep looking

Skills are one crate of 326,149. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.