All skills
aktsmm avatar

/opportunity-factory

@e63d714
by yamapanaktsmm/agent-skills26 stars
4

Run a reusable opportunity-to-artifact workflow: discover unmet needs, set up workspace factories, schedule recurring commander/worker/reporter prompts, batch-refine many items, optionally use SQLite state, build small artifacts, review quality, and track outcomes. Use when the user wants to repeatedly create apps, games, products, content, or experiments from market/user needs.

Use this Skill: https://skilld.dev/gh/aktsmm/agent-skills/opportunity-factory

This session only. Nothing lands on disk.

assetsexamplessetup-packets.md

≈1.7k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Setup Packet Examples

Use these examples as shapes, not product guarantees. Verify current product capabilities before marking any setup unattended.

VS Code GitHub Copilot Chat Supervised Setup

## Opportunity Factory Workspace Setup

- Primary surface: VS Code GitHub Copilot Chat
- Secondary surfaces: none
- Skill location: workspace skill folder or attached skill context
- Prompt runner: manual Agent mode prompt invocation
- Schedule mechanism: manual supervised loop
- State store: workspace-local `factory-state/` folder or user-provided equivalent
- Approval policy: user approval for terminal, write, publish, account, secret, payment, and personal-data actions
- Verification source: VS Code product UI or current docs checked by user/agent
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: supervised
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander -> worker -> commander/reporter-learner

GitHub Copilot CLI Scheduled Setup

## Opportunity Factory Workspace Setup

- Primary surface: GitHub Copilot CLI
- Secondary surfaces: cron / Windows Task Scheduler / CI scheduler
- Skill location: workspace path passed to the CLI or personal skill location if supported
- Prompt runner: CLI wrapper that invokes `assets/prompts/*.md`
- Schedule mechanism: external scheduler
- State store: workspace-local state folder committed or ignored according to repo policy
- Approval policy: no publish/payment/account/secret/personal-data action without explicit approval
- Verification source: CLI help output, current docs, or smoke-tested wrapper command
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: conservative or standard
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander hourly, worker every 2h, reporter daily

GitHub Copilot App Hosted Setup

## Opportunity Factory Workspace Setup

- Primary surface: GitHub Copilot App
- Secondary surfaces: GitHub issues, pull requests, or repo artifacts for state/reporting if supported
- Skill location: repository skill folder or product-supported context attachment
- Prompt runner: product-supported hosted agent prompt/job mechanism, if available in the current version
- Schedule mechanism: verified product scheduler; otherwise use manual or external scheduler
- State store: repository files, issues/discussions, artifacts, or product memory confirmed by current UI/docs
- Approval policy: product-native approval for PRs, file writes, publish, secrets, payment, account, and personal-data actions
- Verification source: current product UI/docs confirming scheduling, persistence, and approval behavior
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: conservative until persistence and approvals pass preflight
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander -> worker -> reporter-learner, with no unattended publish/payment/secret actions

Microsoft Scout Background Setup

## Opportunity Factory Workspace Setup

- Primary surface: Microsoft Scout
- Secondary surfaces: VS Code or repo files for review
- Skill location: Scout-supported workspace or custom skill location, if available in the current version
- Prompt runner: Scout automation / heartbeat / manual chat, depending on verified capabilities
- Schedule mechanism: verified Scout background mode or manual supervised loop
- State store: workspace-local files inside Scout's permitted workspace
- Approval policy: Scout permission gates; stricter policy for background modes than interactive runs
- Verification source: current Scout UI/docs confirming enabled capabilities and permissions
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: conservative first, standard only after one successful preflight
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander -> worker -> reporter-learner, all within permitted workspace scope

Copilot Scheduler VS Code Extension Setup

## Opportunity Factory Workspace Setup

- Primary surface: Copilot Scheduler (VS Code extension)
- Secondary surfaces: VS Code GitHub Copilot Chat for supervised review
- Skill location: workspace skill folder or prompt context used by scheduled tasks
- Prompt runner: Copilot Scheduler scheduled prompt task
- Schedule mechanism: cron expressions in Copilot Scheduler, workspace task scope
- State store: workspace-local `factory-state/` folder or selected SQLite state store
- Approval policy: no unattended publish/payment/account/secret/personal-data/destructive actions; rely on Copilot Chat and VS Code approval prompts where applicable
- Verification source: Copilot Scheduler Marketplace/GitHub README plus current extension UI/settings
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: conservative first, standard only after preflight and one Run Now check pass
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander hourly, worker every 2h with max-runs/day, reporter daily or twice daily

OpenClaw Cron Setup

## Opportunity Factory Workspace Setup

- Primary surface: OpenClaw
- Secondary surfaces: VS Code Chat for manual review
- Skill location: workspace skill folder or runtime-distributed instruction assets
- Prompt runner: OpenClaw agent prompts mapped to commander, worker, reporter-learner
- Schedule mechanism: OpenClaw cron isolated sessions
- State store: shared state folder selected for the factory
- Approval policy: no external publish/payment/account/secret/personal-data action from workers
- Verification source: OpenClaw config/cron status and one successful dry-run pulse
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: conservative first, standard only after preflight passes
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander every 15-30m, workers every 30-60m, reporter every 6-8h

GitHub Actions Low-Frequency Setup

## Opportunity Factory Workspace Setup

- Primary surface: GitHub Actions
- Secondary surfaces: GitHub issues/artifacts for reports
- Skill location: repository files or workflow-provided context
- Prompt runner: workflow command or action that can invoke the chosen agent runner
- Schedule mechanism: low-frequency `schedule` plus `workflow_dispatch`
- State store: repository files, workflow artifacts, issue comments, or external storage approved for the repo
- Approval policy: no external publish/payment/account/secret-dependent action by default
- Verification source: workflow dry run, runner permissions, and repo policy check
- Verification timestamp: <timestamp>
- Checked by: <name or agent>
- Runtime profile: low-cost
- Preflight result: PASS / FAIL / BLOCKED
- First three jobs: commander 1-3h, worker daily/manual, reporter daily

Source: SKILL.md on GitHub

2 warnings22d3 checks · Risk MEDIUM
  • Gen Agent Trust Hub22d

    This skill is an autonomous agent framework that manages complex workflows. It is rated as medium severity because it is self-modifying (the AI edits its own instructions), establishes persistent automated loops, and automatically accesses local secrets for its operations. While it includes many internal safety rules, the high degree of autonomy requires review before deployment.

  • Socket22d

    No alerts

  • Snyk22d

    Risk: MEDIUM · 1 issue

Signed by skilld at e63d714. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 18 hours ago.

Activeupdated 4 weeks ago
argument-hint
対象ドメイン、成果物タイプ、制約、今回の入力
user-invocable
true
metadata
{
  "author": "yamapan (https://github.com/aktsmm)"
}

README badge

README badge for aktsmm/agent-skills/opportunity-factory