wryenmeek/knowledgebase

audit-knowledgebase-workspace

Audits the framework workspace for reference drift, stale commands, attached-tool resolution gaps, and instruction-locality drift. Use when verifying that skills, agents, tests, thin wrappers, and read-only improve dry-runs still point at real repo-local…

原文の言語: 英語

log-intake-rejection

Persists a rejection record in raw/rejected/ when source intake fails. Use when source-intake-steward or evidence-verifier rejects a source and the rejection must be recorded for organizational memory. Write-once, sha256-keyed, locked.

原文の言語: 英語

validate-wiki-governance

Validates wiki governance gates with fixed read-only checks. Use when reviewing wiki safety, proving governance compliance, or before any state-changing wiki action.

原文の言語: 英語

run-deterministic-validators

Runs fixed deterministic validators from the allowlisted knowledgebase execution surface. Use when you need a typed, read-only validator sequence without opening broader automation paths.

原文の言語: 英語

sync-knowledgebase-state

Synchronizes deterministic knowledgebase state through narrow allowlisted wrappers. Use when checking or refreshing wiki/index.md without expanding beyond MVP script boundaries.

原文の言語: 英語

write-sourceref-citations

Writes canonical SourceRef citations from committed repository artifacts. Use when a deterministic helper needs an authoritative provenance string for a raw artifact.

原文の言語: 英語

suggest-backlinks

Recommends governed backlink opportunities between already-cleared pages. Use when editorial follow-up needs contract-aligned related-link suggestions without silently editing the graph.

原文の言語: 英語

extract-entities-and-claims

Extracts candidate entities, concepts, claims, and chronology from a policy-cleared evidence package without opening a write path. Use when a governed synthesis workflow needs a deterministic extraction bundle before drafting or escalation.

原文の言語: 英語

scan-content-freshness

Scans repo-local markdown freshness through the approved validation script family. Use when you need deterministic age checks over wiki or docs content without re-implementing freshness logic inside a skill wrapper.

原文の言語: 英語

synthesize-entity-page

Drafts a policy-cleared entity wiki page from a verified evidence package. Use when evidence-verifier and policy-arbiter have cleared a source intake for entity synthesis, and the synthesis-curator persona must produce a schema-aligned page draft.

原文の言語: 英語

using-agent-skills

Discovers and invokes agent skills. Use when starting a session or when you need to discover which skill applies to the current task. This is the meta-skill that governs how all other skills are discovered and invoked.

原文の言語: 英語

synthesize-concept-page

Drafts a policy-cleared concept wiki page from a verified evidence package. Use when evidence-verifier and policy-arbiter have cleared a source intake for concept synthesis, and the synthesis-curator persona must produce a schema-aligned page draft.

原文の言語: 英語

verified-research

Produces research reports with built-in adversarial verification of every claim before inclusion. Use when conducting comparative analysis, codebase research, or any investigation where accuracy matters more than speed.

原文の言語: 英語

ci-cd-and-automation

Automates CI/CD pipeline setup. Use when setting up or modifying build and deployment pipelines. Use when you need to automate quality gates, configure test runners in CI, or establish deployment strategies.

原文の言語: 英語

code-review-and-quality

Conducts multi-axis code review. Use before merging any change. Use when reviewing code written by yourself, another agent, or a human. Use when you need to assess code quality across multiple dimensions before it enters the main branch.

原文の言語: 英語

performance-optimization

Optimizes application performance. Use when performance requirements exist, when you suspect performance regressions, or when Core Web Vitals or load times need improvement. Use when profiling reveals bottlenecks that need fixing.

原文の言語: 英語

security-and-hardening

Hardens code against vulnerabilities. Use when handling user input, authentication, data storage, or external integrations. Use when building any feature that accepts untrusted data, manages user sessions, or interacts with third-party services.

原文の言語: 英語

grill-me

Stress-tests a plan, spec, or decision through relentless one-at-a-time questioning. Use when validating assumptions, when a plan needs adversarial review, or before committing to a spec. Walks every decision branch until all are resolved.

原文の言語: 英語

idea-refine

Refines ideas iteratively. Refine ideas through structured divergent and convergent thinking. Use "idea-refine" or "ideate" to trigger.

原文の言語: 英語

context-engineering

Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and context for a project.

原文の言語: 英語

caveman

Compresses agent-to-agent output by dropping filler, articles, and pleasantries while keeping full technical accuracy. Use when writing internal agent communications, rubber-duck reports, or sub-agent prompts. NOT for human-facing output.

原文の言語: 英語

edit-article

Improves prose clarity and structure of wiki pages without altering factual content or citations. Use when content is factually correct but reads as AI-generated or is unclear, after detect-ai-tells and enforce-npov have passed. Prose-only: restructure,…

原文の言語: 英語

improve-codebase-architecture

Identifies architecture improvement opportunities and produces structured proposals. Use when reviewing system design, spotting structural issues, or planning architecture evolution.

原文の言語: 英語

planning-and-task-breakdown

Breaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible.

原文の言語: 英語

request-refactor-plan

Creates a structured refactoring proposal as a GitHub Issue. Use when you identify code that needs refactoring and want to document the proposal with motivation, scope, and acceptance criteria.

原文の言語: 英語

triage-issue

Classifies, labels, and prioritizes GitHub Issues with a structured workflow. Use when a new issue needs triage or when reviewing the issue backlog.

原文の言語: 英語

write-a-skill

Guides creation of new agent skills with correct structure, wiring, and tests. Use when creating a new skill for the .github/skills/ framework.

原文の言語: 英語

reconsider-rejected-source

Reconsiders a previously rejected source by setting reconsidered_date and re-entering the intake pipeline. Use when an operator believes a prior rejection should be revisited because circumstances have changed.

原文の言語: 英語

route-wiki-task

Selects the correct governed lane for a wiki work item and hands it off to the appropriate persona. Use when knowledgebase-orchestrator must classify incoming work and route it to the ingest lane, query lane, maintenance lane, or review lane.

原文の言語: 英語

validate-inbox-source

Validates raw inbox material against the repository ingest boundary. Use when checking whether a candidate source can enter the governed intake lane before provenance registration or synthesis.

原文の言語: 英語

enforce-repository-boundaries

Enforces deterministic repository path boundaries for knowledgebase work. Use when validating whether a read or write path stays inside the allowlisted MVP surface.

原文の言語: 英語

patrol-human-edits

Reviews recent human or agent edits to wiki pages for policy, provenance, and citation risk before any lane downstream of change-patrol can open. Use when change-patrol needs a diff-based risk classification for changed wiki content.

原文の言語: 英語

plan-wiki-job

Produces a structured execution plan for a multi-step wiki curation job before any lane work begins. Use when knowledgebase-orchestrator must serialize a complex work sequence, estimate prerequisites, and confirm that all required inputs exist before…

原文の言語: 英語

quality-pass-chain

Runs the post-implementation quality pass in the correct order. Use when completing a non-trivial implementation before merge. Four-step development gate: code-review-and-quality → code-simplification → test-driven-development → documentation-and-adrs.…

原文の言語: 英語

zoom-out

Maps the relevant modules, callers, and abstractions around unfamiliar code or wiki areas. Use when entering unfamiliar territory, when you need to understand how components connect, or when context is unclear before diving into implementation.

原文の言語: 英語

enforce-npov

Applies neutral-point-of-view and attribution policy to evidence-backed wiki work. Use when reviewing a draft, plan, or synthesis for due weight, unsupported inference, or contradiction handling.

原文の言語: 英語

semantic-wiki-lint

Audits wiki pages for semantic maintenance risk including stale summaries, orphaned evidence, and broken structural relationships. Use when maintenance-auditor needs a deterministic evidence surface before recommending any remediation action.

原文の言語: 英語

incremental-implementation

Delivers changes incrementally. Use when implementing any feature or change that touches more than one file. Use when you're about to write a large amount of code at once, or when a task feels too big to land in one step.

原文の言語: 英語

analyze-missed-queries

Scans wiki pages for coverage gaps — missing citations, low-confidence markers, placeholder text, and empty sources — to identify where synthesis work is needed. Use when quality-analyst needs evidence of knowledge gaps before prioritizing new synthesis work.

原文の言語: 英語

append-log-entry

Appends deterministic state-change entries to wiki/log.md. Use when a write-capable flow needs the append-only provenance log after a real state change.

原文の言語: 英語