Yaroslav Berezovskiy
Community@escogido
Yaroslav Berezovskiy publishes read-only review, digest, documentation-audit, and task-selection skills for disciplined task finalization and repository knowledge integrity.
Agent Skills by Yaroslav Berezovskiy
Showing 4 vetted skills indexed across 1 GitHub repositories.
escogidoknowledge-digester
Drafts a read-only digest of durable knowledge captures, stale docs, and contradicted norms from a task delta.
escogidoadversarial-reviewer
Reviews task work read-only for finalization readiness, fidelity, and validation sufficiency.
escogidotask-selection-advisor
Ranks documented task candidates into read-only next-work recommendations in three advisory modes.
escogidodocumentation-auditor
Audits AI-context documentation against code, tests, and schemas to detect stale or contradictory claims.
Frequently Asked Questions About Yaroslav Berezovskiy
FAQPage SchemaWhat tasks can I accomplish with escogido's skills?▼
You can audit documentation against repository evidence, run adversarial pre-finalization reviews of task fidelity and safety, digest accumulated captures and stale docs into a finalization-ready summary, and get ranked read-only advice on which documented task candidate to work on next.
Who are these skills designed for?▼
They target developers and maintainers using a cairn-style task finalization process who need disciplined review gates: documentation auditors verifying docs against code, reviewers checking scope drift and validation sufficiency, and contributors deciding what to work on next.
How are these skills invoked at runtime?▼
Each skill is triggered by explicit user requests or lifecycle transitions: `cairn:digest` or Finalization escalation for knowledge-digester, `cairn:review` for adversarial-reviewer, direct audit requests for documentation-auditor, and 'what next' queries for task-selection-advisor.
Do these skills modify my repository files?▼
No. All four skills are explicitly read-only or advisory: adversarial-reviewer does not edit files or implement fixes, task-selection-advisor must not select or write, knowledge-digester returns a fresh-context digest, and documentation-auditor only investigates and reports discrepancies.
What prerequisites do these skills require?▼
They assume a repository with documented task candidates, AI-context documentation, and a Finalization workflow. documentation-auditor requires accessible code, tests, scripts, or schemas as source-of-truth evidence; knowledge-digester requires an accumulated standing-draft batch of captures.