athola/claude-night-market

这个仓库里有 227 个技能，GitHub 星标 ★ 337。

• claude-code-plugin-referenceExplain plugin, skill, command, agent, and hook mechanics used here. Use when au
• night-market-architecture-contractStates load-bearing decisions, invariants, and weak points. Use when judging a d
• night-market-build-and-envRebuild the dev environment: uv, Python tiers, pins, traps. Use when onboarding
• night-market-change-controlClassify, gate, and review changes. Use when landing a PR, releasing, or amendin
• night-market-collective-memorySearch and record project memory (Discussions, journal, ADRs). Use before re-inv
• night-market-completion-integrity-campaignBind loop 'done' to unfakeable gates. Use to harden egregore/herald loops or pro
• night-market-config-catalogCatalog every config axis, its defaults and guards. Use when adding or auditing
• night-market-debugging-playbookTriage night-market failures by symptom (hooks, CI, tests). Use when a check fai
• night-market-diagnostics-toolkitRun and interpret repo diagnostic scripts (ratchets, validators, token stats). U
• night-market-docs-and-writingMaintain docs of record, ADRs, changelog, and house style. Use when writing repo
• night-market-failure-archaeologyChronicles settled battles, reverts, and dead ends. Use when a fix echoes a past
• night-market-model-and-harness-updatesSweep plugins, skills, agents, commands, and hooks after a model release or Clau
• night-market-operationsRun and ship this repo: make targets, artifacts, release runbook. Use when testi
• night-market-research-frontierMap open problems where this repo can advance SOTA. Use when scoping research. D
• night-market-research-methodologyTurn hunches into accepted results: worthiness score, evidence bar, research-to-
• night-market-validation-and-qaEnforce evidence bar, coverage gates, and regression guards. Use when adding tes
• escalation-governanceAssess whether to escalate models. Use when evaluating reasoning depth.
• friction-detectorDetect friction signals; graduate patterns into rules. Use for session retrospec
• hook-authoringGuide creating Claude Code hooks with security-first design. Use for validation
• hook-scope-guideSelect hook scope (plugin, project, global) by audience. Use when authoring a ho
• hooks-evalEvaluate hook security, performance, and SDK compliance. Use for audits.
• metacognitive-self-modAnalyze and improve the improvement process. Use for detecting regressions and m
• methodology-curatorSurface expert frameworks. Use when creating or evaluating skills, hooks, or age
• modular-skillsBuild composable skill modules with hub-and-spoke loading. Use when token budget
• plugin-reviewReview plugin quality with tiered checks and dependency scoping. Use for PR and
• rules-evalEvaluate Claude Code rules in .claude/rules/. Use for frontmatter, globs, and qu
• shared-patternsProvide reusable patterns for validation, error handling, scaffolding. Use for s
• skill-authoringGuide creating Claude Code skills with TDD and persuasion principles. Use for ne
• skill-graph-auditAudit Skill() refs; detect hubs, isolates, and dangling targets. Use when auditi
• skills-evalEvaluate Claude skill quality through auditing. Use when reviewing or auditing s
• subagent-testingTest skills via TDD in fresh subagents. Use when validating behavior or preventi
• architecture-paradigm-client-serverApplies client-server architecture for web/mobile apps. Use when designing syste
• architecture-paradigm-cqrs-esApplies CQRS and Event Sourcing for read/write separation and audit trails. Use
• architecture-paradigm-domain-drivenModels a business in its own language. Use when the domain has real business rul
• architecture-paradigm-event-drivenApplies event-driven async messaging to decouple producers and consumers. Use wh
• architecture-paradigm-functional-coreApplies Functional Core, Imperative Shell to isolate logic from side effects. Us
• architecture-paradigm-hexagonalApplies hexagonal architecture isolating domain from infrastructure. Use when de
• architecture-paradigm-layeredApplies layered n-tier architecture with enforced boundaries. Use when designing
• architecture-paradigm-microkernelApplies microkernel architecture with minimal core and plugin extensibility. Use
• architecture-paradigm-microservicesApplies microservices for independent deployment and per-service scaling. Use wh
• architecture-paradigm-modular-monolithApplies modular monolith with enforced internal boundaries. Use when teams want
• architecture-paradigm-pipelineApplies pipes-and-filters for sequential data transformations. Use when data flo
• architecture-paradigm-serverlessApplies serverless FaaS patterns for event-driven workloads. Use when designing
• architecture-paradigm-service-basedApplies coarse-grained service architecture for deployment independence. Use whe
• architecture-paradigm-space-basedApplies data-grid architecture for high-traffic stateful workloads. Use when a s
• architecture-paradigmsSelects and routes to the right architecture paradigm. Use when choosing pattern
• architecture-aware-initSelects architecture paradigm via research before scaffolding. Use when architec
• dorodangoPolishes working code through successive quality passes in fresh subagents. Use
• makefile-generationGenerates Makefiles with testing, linting, formatting, and automation targets. U
• mission-orchestratorOrchestrates full project lifecycle by auto-detecting state and routing to the c
• precommit-setupConfigures pre-commit hooks for linting, type checking, formatting, and testing.
• project-brainstormingGuides project ideation via Socratic questioning to produce a validated brief. U
• project-executionExecutes implementation plans with progress tracking, checkpoint validation, and
• project-initScaffolds new projects with git, CI/CD workflows, pre-commit hooks, and build co
• project-planningConverts a specification into a phased, dependency-ordered implementation plan.
• project-specificationTransforms project briefs into testable specifications with user stories and acc
• skill-library-missionBuild a project skill library in .claude/skills/ via discovery, parallel authori
• war-roomConvenes a multi-LLM expert panel to pressure-test hard-to-reverse decisions. Us
• war-room-checkpointAssesses decision reversibility and risk at critical checkpoints. Use when a wor
• workflow-setupConfigures GitHub Actions CI/CD workflows for testing, linting, and deployment.
• architecture-diagramGenerates a Mermaid architecture diagram showing high-level component relationsh
• call-chainTraces execution paths through the code graph with criticality scoring and Merma
• class-diagramGenerates a Mermaid class diagram showing types, inheritance, and composition. U
• code-communitiesDetects architectural clusters and coupling boundaries via community detection o
• data-flowGenerates a Mermaid sequence diagram showing how data moves between components.
• dependency-graphGenerates a Mermaid dependency graph showing import relationships between module
• workflow-diagramGenerates a Mermaid workflow diagram showing process steps, decisions, and state
• agent-teamsCoordinates Claude agent teams via filesystem protocol. Use when orchestrating p
• codex-delegationDelegates tasks to the OpenAI Codex CLI (codex exec) via delegation-core. Use wh
• delegation-coreDelegates execution to eight CLIs (Gemini, Qwen, MiniMax, GLM, Muse, Codex, Open
• gemini-delegationDelegates tasks to Gemini CLI implementing delegation-core for Google's models.
• glimmer-delegationDelegates tasks to a locally served Muse Glimmer via ollama. Use when delegation
• glm-delegationDelegates tasks to Z.ai GLM-5.x via the stock claude binary and an endpoint swap
• minimax-delegationDelegates tasks to the MiniMax CLI (mmx) via delegation-core. Use when delegatio
• muse-delegationDelegates tasks to Meta's Muse Code CLI (muse) via delegation-core. Use when del
• opencode-delegationDelegates tasks to OpenCode (opencode run) via delegation-core. Use when delegat
• provider-setupReports which delegation CLIs are installed and authenticated, installs missing
• qwen-delegationDelegates tasks to Qwen CLI via delegation-core for Alibaba's models. Use when d
• action-first-outputShapes turns action-first: action leads, steps numbered, state restated. Use for
• agent-expenditureTracks per-agent token usage and flags waste in parallel dispatch. Use when eval
• bloat-detectorDetects codebase bloat via dead code, duplication, complexity, and doc bloat sca
• clear-contextManages context overflow by handing off to a fresh subagent at 80% usage. Use wh
• code-quality-principlesApplies KISS, YAGNI, and SOLID principles for clean code with reduced complexity
• compression-strategyRecommends context compression strategies for bloated or quota-heavy sessions. U
• context-mapGenerates a compressed project context map to avoid expensive Read/Grep calls. U
• context-optimizationOptimizes context window via MECW principles and memory tiering. Use when contex
• cpu-gpu-performanceEstablishes CPU/GPU baselines before resource-intensive operations. Use before b
• decisive-actionGuides when to ask clarifying questions versus proceed autonomously. Use to redu
• elegant-codeGuide minimal code via a decision ladder with full safety, edge, and negative-ca
• mcp-code-executionRoutes multi-tool workflows through MCP servers for large datasets and pipelines
• response-compressionCompresses verbose responses by removing filler and framing to save 200-400 toke
• smart-sourcingSelects optimal sources for tool calls, balancing accuracy with token cost. Use
• token-conservationEnforces token quota management at session start with conservation and compressi
• install-watchdogInstalls egregore watchdog daemon via launchd or systemd for autonomous relaunch
• quality-gateOrchestrates the QUALITY pipeline stage for egregore work items, running code re
• summonAutonomous orchestrator for manifest work items through the development lifecycl
• uninstall-watchdogRemoves the egregore watchdog daemon and its associated files. Use when stopping
• challengePresents adaptive codebase challenge questions with multiple-choice and trace ex
• curateAdds developer-authored annotations to the gauntlet knowledge base. Use when cap
• extractBuilds the gauntlet knowledge base from AST extraction and AI enrichment. Use wh
• gauntlet-curateAudits the DSA problem bank for coverage gaps and proposes new YAML entries. Use
• graph-buildBuilds or updates the code knowledge graph via tree-sitter AST and SQLite. Use w
• graph-searchSearches the code knowledge graph by function, class, or type using FTS5 full-te
• onboardGuides a new developer through five staged challenge sets covering architecture,
• rule-catalogBrowse hookify rule catalog. Use when installing pre-built rules or browsing cat
• writing-rulesCreates behavioral rules in markdown to block dangerous commands or restrict AI
• assisted-masteryMakes agent reasoning visible, surfaces tradeoffs, and fades help so humans buil
• catchupSummarizes recent git changes for context recovery after session breaks. Use whe
• dependency-verificationVerifies a package exists before install, defending against hallucination and sl
• diff-analysisAnalyzes changesets with risk scoring, categorization by type and impact, and re
• feature-reviewScores backlog items with RICE/WSJF/Kano and files GitHub issues for top candida
• graduated-implementationRamps implementation ambition a notch only after the prior increment is understo
• justifyAudits changes for additive bias and Iron Law compliance. Use when reviewing com
• karpathy-principlesPre-implementation gate covering think-first, simplicity, surgical edits, and ve
• latent-space-engineeringShapes agent behavior via instruction framing and style transfer. Use when compo
• proof-of-workEnforces validation and evidence before claiming work complete. Use before decla
• review-coreProvides review-workflow scaffolding for context, evidence, and output. Use at t
• rigorous-reasoningApplies anti-sycophancy checklist to override agreement bias. Use when analyzing
• scope-guardScores feature worthiness and enforces branch-size limits against overengineerin
• structured-outputFormats review deliverables with consistent structure for comparable findings. U
• vow-enforcementClassifies and enforces constraints via soft vows, hard vows, and Nen Court laye
• workflow-monitorDetects workflow failures and inefficient patterns then files GitHub issues. Use
• additive-bias-defenseInverts burden of proof for code additions. Use when reviewing PRs, planning ref
• authentication-patternsProvides auth patterns for API keys, OAuth, and token management. Use when imple
• content-sanitizationProvides sanitization guidelines for external content in skills and hooks. Use w
• damage-controlRecovers broken agent state via crash recovery, context overflow, and merge conf
• decision-journalContract for the project decision journal (tradeoffs and lessons-learned logs).
• deferred-captureDefines the contract for deferred-item capture across plugins. Use when building
• document-conversionConverts documents and URLs to markdown via tiered fallback (MCP markitdown, nat
• error-patternsProvides error classification, recovery, and graceful-degradation patterns. Use
• evaluation-frameworkProvides weighted scoring, rubrics, and decision-threshold patterns. Use when de
• git-platformDetects git forge (GitHub/GitLab/Bitbucket) and maps CLI commands cross-platform
• loop-optimizationDecides hand-vs-compiler for loop transforms (unrolling, SIMD, fusion, branchles
• markdown-formattingEnforces markdown line-wrap and structure rules for clean git diffs. Use when wr
• progressive-loadingImplements hub-and-spoke lazy loading to minimize token usage in large skills. U
• pytest-configProvides standardized pytest config, reusable fixtures, and CI integration patte
• quota-managementTracks quotas, monitors thresholds, and degrades gracefully for rate-limited API
• risk-classificationClassifies agent tasks into 4 risk tiers (GREEN/YELLOW/RED/CRITICAL). Use when a
• sem-integrationProvides sem semantic-diff detection, install-on-first-use, and fallback pattern
• service-registryRegisters external services with health checks, central config, and unified exec
• stewardshipApplies stewardship virtues (Care, Curiosity, Humility, Diligence) to plugin wor
• storage-templatesProvides templates and lifecycle patterns for storage and documentation systems.
• supply-chain-advisoryAudits dependency supply chains for bad versions, lockfile drift, and artifact i
• testing-quality-standardsDefines testing quality metrics, coverage thresholds, and anti-patterns. Use whe
• usage-loggingImplements structured usage logging and audit trails for cost and session tracki
• utilityScores agent actions by expected gain, cost, uncertainty, and redundancy. Use wh
• digital-garden-cultivatorManages digital garden notes, link structures, and health metrics. Use when cura
• knowledge-intakeProcesses external resources into stored knowledge with quality scoring and rout
• knowledge-locatorSearches and navigates stored knowledge in memory palaces. Use when looking for
• memory-clarity-probe>
• memory-palace-architectDesigns memory palace structures with spatial layouts and domain organization. U
• palace-diagramGenerates Mermaid and ASCII diagrams of palace structure, knowledge topology, an
• palace-index-curatorCurate the web-capture index. Use when the capture backlog grows, captures sit u
• review-chamberCaptures and retrieves PR-review findings in memory palaces. Use after PR review
• session-handoffDecompose a session into typed handoff units and recall them later. Use when end
• session-palace-builderBuilds session-scoped temporary memory palaces for extended conversations. Use w
• dora-metricsComputes DORA delivery-performance metrics from git and GitHub API. Use when ass
• github-initiative-pulseGenerates markdown digests and CSV exports for GitHub initiative health. Use whe
• release-health-gatesStandardizes release approvals with GitHub-aware checklists and deployment gates
• setupProvisions the oracle ML inference daemon with onnxruntime via uv. Use when sett
• python-asyncAsync Python patterns via asyncio and aiohttp for I/O-bound concurrency. Use whe
• python-packagingPython package creation and PyPI distribution via pyproject.toml and entry point
• python-performanceProfiles Python code for performance bottlenecks and memory issues. Use when Pyt
• python-testingPython testing patterns with pytest, fixtures, TDD, mocking, async and integrati
• api-reviewEvaluates API surface design, consistency, and exemplar alignment. Use when revi
• architecture-reviewAssesses architecture decisions, ADR compliance, and coupling. Use when evaluati
• blast-radiusAnalyzes code change impact with risk scoring and affected-node mapping. Use bef
• bug-reviewHunts bugs with evidence trails. Use when investigating unexpected behavior or b
• code-refinementImproves code quality across duplication, efficiency, and architectural fit. Use
• hardenApplies NIST/CWE security hardening to Python and Rust code. Use when auditing c
• makefile-reviewAudits Makefiles for build correctness, portability, and recipe duplication. Use
• math-reviewVerifies math-heavy code for algorithmic correctness and numerical stability. Us
• performance-reviewDetects time and space complexity hotspots via AST scan. Use when code feels slo
• rust-reviewAudits Rust code for unsafe blocks, ownership issues, and Cargo dependency risks
• safety-critical-patternsApplies NASA Power of 10 rules for safety-critical verifiable code. Use when aud
• shell-reviewAudits shell scripts for correctness, portability, and common pitfalls. Use when
• test-reviewEvaluates test suites for coverage gaps, TDD/BDD compliance, and anti-patterns.
• tiered-auditRuns a three-tier codebase audit (git history, targeted scans, full review) with
• unified-reviewOrchestrates multi-domain review (code, arch, tests, security) in a single pass.
• computer-controlAutomates desktop GUI workflows via computer use API with screenshot capture. Us
• commit-messagesGenerates conventional commit messages from staged changes. Use when committing
• do-issueImplements GitHub or GitLab issues via parallel subagents with review gates betw
• doc-consolidationMerges ephemeral report and analysis artifacts into permanent documentation. Use
• doc-updatesUpdates documentation after code changes with quality gates, slop detection, and
• file-analysisMaps file structure and module organization of a codebase. Use before architectu
• git-workspace-reviewVerifies workspace state and staged changes as a read-only preflight. Use before
• pr-prepPrepares pull requests by running quality gates, drafting descriptions, and vali
• pr-reviewReviews pull requests with scope validation, requirements compliance, and line c
• session-managementManages Claude Code sessions with naming, checkpointing, and resume strategies.
• stack-createInitializes a stacked branch set from an ordered plan, one branch per slice with
• stack-modeDetects shared stack membership and iterates a command across all PRs in base-to
• stack-pushPushes all branches in a stack and opens or updates one dependent PR per slice.
• stack-rebaseCascades a rebase through an entire PR stack after a base PR merges or upstream
• test-updatesUpdates, generates, and validates tests using git-workspace context and TDD/BDD
• tutorial-updatesGenerates or updates tutorials from VHS tapes and Playwright specs with dual-ton
• update-readmeRefreshes README structure favoring concision (an index, not a manual). Use when
• validate-prGenerates and self-executes a diff-derived test plan for a PR.
• version-updatesBumps versions, updates changelogs, and coordinates version changes across files
• workflow-improvementEvaluates and improves skills, agents, commands, and hooks after a workflow slic
• doc-generatorGenerates or remediates documentation with human-quality writing. Use when creat
• doc-importerConverts external documents (PDF, DOCX, PPTX, XLSX, HTML) into editable markdown
• session-replayConverts a Claude Code session JSONL file into an animated GIF terminal replay.
• session-to-postConverts a Claude Code session into a blog post, case study, or Reddit post. Use
• simplified-technical-englishApplies an ASD-STE100-derived register to operator and procedural text. Use when
• slop-detectorDetects AI-generated writing patterns in prose. Use when reviewing docs for slop
• style-learnerExtracts writing style patterns from exemplar text into a reusable profile. Use
• tech-tutorialPlans, drafts, and refines technical tutorials for developers. Use when writing
• voice-extractExtracts a user's writing voice from text samples via SICO comparative analysis.
• voice-generateGenerates text in a learned writing voice. Use when drafting content that must m
• voice-learnImproves a voice profile by learning from manual edits. Use after editing genera
• voice-reviewRuns parallel prose and craft review agents against a voice profile. Use when ch
• browser-recordingRecords browser sessions via Playwright and converts video to GIF. Use when crea
• gif-generationConverts webm/mp4 video files to optimized GIFs via ffmpeg with configurable qua
• media-compositionCombines GIFs and videos into composite tutorials with vertical or grid layouts
• vhs-recordingGenerates terminal recordings using VHS tape scripts and produces GIF outputs. U
• spec-writingCreates clear, testable specifications from feature descriptions with user stori
• speckit-orchestratorOrchestrates Spec Driven Development by coordinating spec, plan, and task skills
• task-planningGenerates phased, dependency-ordered implementation tasks from specifications. U
• code-searchSearches GitHub for existing implementations, libraries, or patterns. Use when f
• digRefines an active research session by drilling deeper into a subtopic. Use after
• discourseScans HN, Lobsters, Reddit, and tech blogs for community experience reports. Use
• exportExport a research session into knowledge-intake compatible markdown. Use after a
• ideateGenerate diverse solution candidates with category-spanning ideation methods and
• papersSearches academic literature via arXiv, Semantic Scholar, and open-access PDFs.
• researchRuns multi-source research across GitHub, HN, Reddit, arXiv, and Semantic Schola
• synthesizeMerges, dedupes, ranks, and formats research findings into a report. Use after r
• trizApplies TRIZ cross-domain analogical reasoning to find solutions from adjacent f