agentsclimarketplace

Apple dash docsets

Skill Syniox5334/apple-dev-skills/skills/apple-dash-docsets

Provide essential skills for Apple development, including Xcode workflows, Dash docset management, and Swift package scaffolding and verification.

Install
npx -y skills add Syniox5334/apple-dev-skills --skill apple-dash-docsets

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

  • 2 stars2 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.

What its author says it does

Copied from the file, not written here

Manage Dash docsets and cheatsheets on macOS. Use when tasks involve Dash search and discovery, installing missing docsets, or generating guidance when a docset is unavailable.

SKILL.md

5.3 KB, as published. Nobody here has run it

Apple Dash Docsets

Purpose

Manage Dash docsets and cheatsheets on macOS with one top-level entry point. scripts/run_workflow.py is the authoritative runtime path for stage selection, fallback order, source priority, approval gating, and structured generation guidance; it does not replace the agent's own Dash access methods.

When To Use

  • Use this skill for Dash search and discovery requests.
  • Use this skill for Dash installation requests after search has identified a missing docset.
  • Use this skill for Dash generation guidance when installation cannot complete.
  • Use this skill when the user needs the exact fallback path between agent-side Dash MCP usage, local HTTP, and URL or Service integration.
  • Recommend apple-xcode-workflow when the user needs Apple or Swift execution, diagnostics, build or test work, or Apple docs reasoning outside Dash management.
  • Recommend apple-swift-package-bootstrap when the user is starting a brand new Swift package rather than managing Dash content.

Single-Path Workflow

  1. Classify the request into one stage:
    • search
    • install
    • generate
  2. If no stage is explicit, start at search.
  3. Run scripts/run_workflow.py with the selected stage:
    • search: applies configured fallback order and returns a structured access-path decision
    • install: applies configured source priority and approval gating
    • generate: returns structured automation-first guidance
  4. If the selected stage cannot complete, hand off forward through references/stage-handoff-contract.md:
    • search -> install
    • install -> generate
  5. Return one status, one path_type, and one output contract for the stage that ran.

Inputs

  • stage: search, install, or generate
  • query: required for search
  • docset_identifiers: optional for search; use installed identifiers only
  • docset_request: required for install and generate
  • approval: required before side-effectful install actions
  • Defaults:
    • runtime entrypoint: python3 scripts/run_workflow.py
    • search access order is mcp -> http -> url-service
    • install source priority is built-in,user-contributed,cheatsheet
    • default search result limit is 20
    • default search snippets setting is true

Outputs

  • status
    • success: the selected stage completed on its primary or fallback path
    • blocked: prerequisites, approval, or usable access paths are missing
    • handoff: the current stage is handing off to the next Dash stage
  • path_type
    • primary: the selected stage completed normally
    • fallback: the selected stage completed through its documented fallback path
  • output
    • stage
    • access_path or source_path
    • matches
    • install result or generation guidance when applicable
    • one next step when follow-up is required

Guards and Stop Conditions

  • Do not run install actions without explicit user approval.
  • Do not invent docset identifiers or catalog matches.
  • Stop with blocked when agent-side Dash MCP usage, HTTP, and URL or Service paths are all unusable for search.
  • Stop with blocked when install or generate lacks a concrete docset request.
  • Keep search, install, and generate in forward stage order; do not blend them into competing primary workflows.

Fallbacks and Handoffs

  • search falls back in this order: agent-side Dash MCP usage, local HTTP API, then URL or Service guidance.
  • install hands off to generate when no installable catalog match exists.
  • generate falls back from stable automation to deterministic manual guidance.
  • Use references/stage-handoff-contract.md when search transitions to install or install transitions to generate.
  • Recommend apple-xcode-workflow directly when the user’s task shifts from docs management to Apple or Swift execution work.
  • Recommend apple-swift-package-bootstrap directly when the user needs new Swift package scaffolding.
  • scripts/run_workflow.py is the only local runtime entrypoint for stage behavior; helper scripts remain implementation details behind it.

Customization

  • Use references/customization-flow.md.
  • scripts/customization_config.py stores and reports customization state.
  • scripts/run_workflow.py loads and enforces the runtime-safe knobs documented in references/customization-flow.md.

References

Workflow References

  • references/dash_mcp_tools.md
  • references/dash_http_api.md
  • references/dash_url_and_service.md

Contract References

  • references/stage-handoff-contract.md
  • references/automation-prompts.md
  • references/customization-flow.md

Support References

  • Recommend references/snippets/apple-swift-core.md when the user needs a reusable Apple and Swift baseline snippet in their own repo alongside Dash-based docs workflows.
  • references/catalog_built_in_docsets.json
  • references/catalog_user_contrib_docsets.json
  • references/catalog_cheatsheets.json
  • references/snippets/apple-swift-core.md

Script Inventory

  • scripts/run_workflow.py
  • scripts/dash_api_probe.py
  • scripts/dash_catalog_match.py
  • scripts/dash_catalog_refresh.py
  • scripts/dash_url_search.py
  • scripts/dash_url_install.py

Keep looking

Skills are one crate of 328,083. 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.