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.

referencessqlite-state-store.md

≈711 tokens on demand. Your agent reads this file only when SKILL.md points to it.

SQLite State Store

SQLite is an optional state backend for large, repeated, or scheduled Opportunity Factory runs. JSON remains the default for small/manual work.

Use SQLite When

  • The factory tracks 100+ items or candidates.
  • Rubber-duck review runs multiple passes per item.
  • Multiple scheduled workers need claim/retry behavior.
  • Resume, dedupe, aggregate reporting, or historical audit queries matter.

Avoid SQLite When

  • A user is doing a small supervised run.
  • The platform cannot persist local database files safely.
  • Repository policy forbids generated binary state files.
  • A hosted platform already provides durable state, locks, and queryable logs.

Schema

Use assets/templates/factory-state.sqlite.sql as the minimal schema. It models:

  • runs: top-level factory executions
  • items: opportunities or products being refined
  • tasks: queue entries
  • claims: worker locks and retries
  • reviews: pass-indexed rubber-duck findings
  • critic_log: durable repair/re-review receipt, finding resolution, validation result, and independence evidence
  • repair_attempts: parent-task reservation and finalization for bounded repair/re-review accounting
  • artifacts: durable outputs
  • outcomes: observed/estimated/assumed metrics
  • pipeline_log: append-only audit events

Initialize

Use scripts/init_factory_sqlite.py when a selected state store should be a SQLite database.

python scripts/init_factory_sqlite.py --db-path <state-dir>/factory-state.sqlite --domain <domain> --audience <audience> --apply

Omit --apply for dry-run. Existing databases are not overwritten unless --force is provided.

Rules

  • Keep worker writes artifact-first when possible; let commander/reducer import rows.
  • Use one writer or transactional writes for state mutation.
  • Store repair child tasks with parent_task_id, finding IDs, input hash, and acceptance checks. Commander inserts repair_attempts.status=reserved before dispatch, then finalizes the same row as validation pass/fail or repair-start-failed; use it to compute the parent task's remaining repair budget. Store validation and re-review receipts in critic_log, including repair task ID, role/layer, immutable adapter receipt source/reference/hash, and constrained independence/next-state values; do not treat a worker's repair output as completion before the independent receipt exists.
  • Keep secrets and personal data out of the database.
  • Do not commit SQLite database files unless repository policy explicitly allows it.
  • Prefer SQL schema in the skill; create actual .sqlite files only in the user's selected state store.

Smoke Test

The validator should execute the schema with Python stdlib sqlite3 against an in-memory database. This proves the schema parses without creating a persistent database.

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